| { | |
| "best_metric": 0.6386890411376953, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-2_120-samples_config-2_auto/checkpoint-44", | |
| "epoch": 14.909090909090908, | |
| "eval_steps": 500, | |
| "global_step": 82, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.7632969617843628, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.045, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.6736323237419128, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9398, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.7394237518310547, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.0219, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 1.0249334573745728, | |
| "eval_runtime": 9.6562, | |
| "eval_samples_per_second": 2.485, | |
| "eval_steps_per_second": 2.485, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.7125771641731262, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.9723, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.7020024657249451, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.9729, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.6056334376335144, | |
| "learning_rate": 4e-05, | |
| "loss": 0.9649, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.8979780673980713, | |
| "eval_runtime": 9.3902, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.4897518754005432, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.8108, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.514333963394165, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.8387, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.5265169143676758, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.7596, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 0.8028402328491211, | |
| "eval_runtime": 9.3936, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.4110087752342224, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.7674, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.3981207311153412, | |
| "learning_rate": 8e-05, | |
| "loss": 0.6913, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.40870991349220276, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.6873, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.7306434512138367, | |
| "eval_runtime": 9.3967, | |
| "eval_samples_per_second": 2.554, | |
| "eval_steps_per_second": 2.554, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.3543305993080139, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.6631, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.36543551087379456, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.6361, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 0.6907404065132141, | |
| "eval_runtime": 9.3877, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.35531890392303467, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.5986, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.3549633324146271, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.5879, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.3520183861255646, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.5688, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.6625128984451294, | |
| "eval_runtime": 9.3875, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.3167888820171356, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.5587, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.30778682231903076, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.5389, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.3280119299888611, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.5268, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 0.6464441418647766, | |
| "eval_runtime": 9.3923, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.3251132667064667, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.5078, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.3697480857372284, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.452, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.34461918473243713, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.4821, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.6386890411376953, | |
| "eval_runtime": 9.3913, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.3459834158420563, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.4233, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.4492061734199524, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.439, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 0.6490373015403748, | |
| "eval_runtime": 9.3854, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.4329462945461273, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.4266, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.4201180040836334, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.3699, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.4516981840133667, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.3766, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.6785171627998352, | |
| "eval_runtime": 9.3932, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.4533138871192932, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.3346, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.5237334370613098, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.3134, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.5634166598320007, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.3045, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 0.7300302982330322, | |
| "eval_runtime": 9.3837, | |
| "eval_samples_per_second": 2.558, | |
| "eval_steps_per_second": 2.558, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.5275758504867554, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.2601, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.6040693521499634, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.2447, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.7197152376174927, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.2261, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.8326237797737122, | |
| "eval_runtime": 9.3887, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.8877094388008118, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.1762, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 0.7513244152069092, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.1684, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 0.9589914679527283, | |
| "eval_runtime": 9.3858, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.6802906394004822, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.1507, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.694140613079071, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.1268, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 0.7362017631530762, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.097, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.1235949993133545, | |
| "eval_runtime": 9.3877, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.181818181818182, | |
| "grad_norm": 0.6568365097045898, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.0815, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 14.545454545454545, | |
| "grad_norm": 0.7499954104423523, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.0688, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "grad_norm": 0.6286192536354065, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.0478, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "eval_loss": 1.2808623313903809, | |
| "eval_runtime": 9.4085, | |
| "eval_samples_per_second": 2.551, | |
| "eval_steps_per_second": 2.551, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "step": 82, | |
| "total_flos": 4.438596877759283e+16, | |
| "train_loss": 0.49398956584130843, | |
| "train_runtime": 1893.2723, | |
| "train_samples_per_second": 2.324, | |
| "train_steps_per_second": 0.132 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.438596877759283e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |