| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.999819135467535, | |
| "eval_steps": 500, | |
| "global_step": 1382, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03617290649303671, | |
| "grad_norm": 5.598344326019287, | |
| "learning_rate": 1.0000000000000001e-07, | |
| "loss": 0.8052, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07234581298607343, | |
| "grad_norm": 15.454778671264648, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 1.3834, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10851871947911014, | |
| "grad_norm": 6.023103713989258, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 0.6989, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.14469162597214685, | |
| "grad_norm": 14.213159561157227, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 0.8097, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18086453246518358, | |
| "grad_norm": 3.366466999053955, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.2367, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.21703743895822028, | |
| "grad_norm": 15.736528396606445, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 0.3249, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.25321034545125704, | |
| "grad_norm": 2.5983786582946777, | |
| "learning_rate": 7.000000000000001e-07, | |
| "loss": 0.1757, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2893832519442937, | |
| "grad_norm": 10.108122825622559, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 0.2398, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.32555615843733043, | |
| "grad_norm": 2.5293455123901367, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 0.1667, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.36172906493036716, | |
| "grad_norm": 9.013643264770508, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.22, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3979019714234039, | |
| "grad_norm": 3.5457277297973633, | |
| "learning_rate": 1.1e-06, | |
| "loss": 0.168, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.43407487791644056, | |
| "grad_norm": 6.44632625579834, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.2021, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4702477844094773, | |
| "grad_norm": 2.5131373405456543, | |
| "learning_rate": 1.3e-06, | |
| "loss": 0.1573, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5064206909025141, | |
| "grad_norm": 10.056652069091797, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 0.1839, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5425935973955507, | |
| "grad_norm": 1.9256174564361572, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.1432, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5787665038885874, | |
| "grad_norm": 5.919682025909424, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.178, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6149394103816241, | |
| "grad_norm": 2.477151393890381, | |
| "learning_rate": 1.7000000000000002e-06, | |
| "loss": 0.1377, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6511123168746609, | |
| "grad_norm": 6.950107097625732, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 0.1865, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6872852233676976, | |
| "grad_norm": 2.580976963043213, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 0.1296, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7234581298607343, | |
| "grad_norm": 6.907202243804932, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.1769, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.759631036353771, | |
| "grad_norm": 2.5415403842926025, | |
| "learning_rate": 2.1000000000000002e-06, | |
| "loss": 0.1345, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.7958039428468078, | |
| "grad_norm": 8.191935539245605, | |
| "learning_rate": 2.2e-06, | |
| "loss": 0.1727, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8319768493398445, | |
| "grad_norm": 4.426870822906494, | |
| "learning_rate": 2.3000000000000004e-06, | |
| "loss": 0.1267, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8681497558328811, | |
| "grad_norm": 5.906033515930176, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.1617, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9043226623259178, | |
| "grad_norm": 2.5345969200134277, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1444, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9404955688189546, | |
| "grad_norm": 5.994137287139893, | |
| "learning_rate": 2.6e-06, | |
| "loss": 0.1766, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9766684753119913, | |
| "grad_norm": 3.034348249435425, | |
| "learning_rate": 2.7000000000000004e-06, | |
| "loss": 0.13, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.9998191354675349, | |
| "eval_loss": 0.1486499309539795, | |
| "eval_runtime": 18037.1186, | |
| "eval_samples_per_second": 0.526, | |
| "eval_steps_per_second": 0.526, | |
| "eval_wer": 0.10366752081998719, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.0128413818050281, | |
| "grad_norm": 2.1960830688476562, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.1467, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0490142882980646, | |
| "grad_norm": 2.957355260848999, | |
| "learning_rate": 2.9e-06, | |
| "loss": 0.1113, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.0851871947911014, | |
| "grad_norm": 1.6470922231674194, | |
| "learning_rate": 3e-06, | |
| "loss": 0.1469, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.121360101284138, | |
| "grad_norm": 3.09590220451355, | |
| "learning_rate": 3.1000000000000004e-06, | |
| "loss": 0.1034, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.1575330077771748, | |
| "grad_norm": 2.3706791400909424, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.1463, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1937059142702116, | |
| "grad_norm": 1.995174765586853, | |
| "learning_rate": 3.3000000000000006e-06, | |
| "loss": 0.1004, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.2298788207632483, | |
| "grad_norm": 1.665616750717163, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.1297, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.266051727256285, | |
| "grad_norm": 2.44356369972229, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.1112, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.3022246337493217, | |
| "grad_norm": 1.8907896280288696, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.1379, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.3383975402423585, | |
| "grad_norm": 2.632768154144287, | |
| "learning_rate": 3.7e-06, | |
| "loss": 0.1011, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.3745704467353952, | |
| "grad_norm": 1.6871304512023926, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.133, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.410743353228432, | |
| "grad_norm": 2.9852051734924316, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 0.1126, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.4469162597214686, | |
| "grad_norm": 2.0160508155822754, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.1361, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.4832700307469704, | |
| "grad_norm": 2.4528961181640625, | |
| "learning_rate": 9.345549738219896e-06, | |
| "loss": 0.1073, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.5194429372400071, | |
| "grad_norm": 3.373425006866455, | |
| "learning_rate": 8.691099476439791e-06, | |
| "loss": 0.1388, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.5556158437330438, | |
| "grad_norm": 2.545165538787842, | |
| "learning_rate": 8.036649214659686e-06, | |
| "loss": 0.1101, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.5917887502260806, | |
| "grad_norm": 1.6243386268615723, | |
| "learning_rate": 7.382198952879581e-06, | |
| "loss": 0.1339, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6279616567191173, | |
| "grad_norm": 3.5856447219848633, | |
| "learning_rate": 6.727748691099477e-06, | |
| "loss": 0.1069, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.664134563212154, | |
| "grad_norm": 1.4985790252685547, | |
| "learning_rate": 6.073298429319372e-06, | |
| "loss": 0.1331, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7003074697051908, | |
| "grad_norm": 3.4537761211395264, | |
| "learning_rate": 5.418848167539268e-06, | |
| "loss": 0.0903, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.7364803761982275, | |
| "grad_norm": 2.7335891723632812, | |
| "learning_rate": 4.764397905759163e-06, | |
| "loss": 0.1207, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.7726532826912642, | |
| "grad_norm": 2.135629892349243, | |
| "learning_rate": 4.109947643979058e-06, | |
| "loss": 0.0948, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.808826189184301, | |
| "grad_norm": 1.6300333738327026, | |
| "learning_rate": 3.455497382198953e-06, | |
| "loss": 0.1099, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.8449990956773377, | |
| "grad_norm": 2.582153797149658, | |
| "learning_rate": 2.8010471204188483e-06, | |
| "loss": 0.0896, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.8811720021703744, | |
| "grad_norm": 1.5254755020141602, | |
| "learning_rate": 2.1465968586387435e-06, | |
| "loss": 0.0942, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9173449086634111, | |
| "grad_norm": 3.2325875759124756, | |
| "learning_rate": 1.4921465968586387e-06, | |
| "loss": 0.0932, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.9535178151564478, | |
| "grad_norm": 1.1424560546875, | |
| "learning_rate": 8.376963350785341e-07, | |
| "loss": 0.1005, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.9896907216494846, | |
| "grad_norm": 3.229748010635376, | |
| "learning_rate": 1.8324607329842932e-07, | |
| "loss": 0.0844, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.999819135467535, | |
| "eval_loss": 0.14780554175376892, | |
| "eval_runtime": 18259.5061, | |
| "eval_samples_per_second": 0.52, | |
| "eval_steps_per_second": 0.52, | |
| "eval_wer": 0.10174567584881486, | |
| "step": 1382 | |
| }, | |
| { | |
| "epoch": 1.999819135467535, | |
| "step": 1382, | |
| "total_flos": 1.5025096510930944e+20, | |
| "train_loss": 0.029662437230219198, | |
| "train_runtime": 33102.9931, | |
| "train_samples_per_second": 1.336, | |
| "train_steps_per_second": 0.042 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 1382, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5025096510930944e+20, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |