| { | |
| "best_metric": 0.7893829345703125, | |
| "best_model_checkpoint": "data/Mistral-7B_task-2_120-samples_config-2_full/checkpoint-49", | |
| "epoch": 16.0, | |
| "eval_steps": 500, | |
| "global_step": 88, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 1.2024271488189697, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.2262, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 1.0684531927108765, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.1611, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 1.0697492361068726, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.2098, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 1.1455994844436646, | |
| "eval_runtime": 30.0121, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.9591270685195923, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.1585, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.7196555733680725, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.1039, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.4609065353870392, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0811, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.0115667581558228, | |
| "eval_runtime": 30.0621, | |
| "eval_samples_per_second": 0.798, | |
| "eval_steps_per_second": 0.798, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.3170148432254791, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.9785, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.3291914165019989, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.9771, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.3540711998939514, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.952, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 0.9449048042297363, | |
| "eval_runtime": 30.0139, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.38774189352989197, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.9335, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.4278547763824463, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8671, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.27088961005210876, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.8442, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.8452054858207703, | |
| "eval_runtime": 30.0171, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.24265070259571075, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.8031, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.23036198318004608, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.7828, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 0.8171373009681702, | |
| "eval_runtime": 30.0154, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.19761155545711517, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.7779, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.2000536471605301, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.7639, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.1745884269475937, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.7505, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.8017528653144836, | |
| "eval_runtime": 30.0131, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.19891045987606049, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.7421, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.17337866127490997, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.7222, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.22037535905838013, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.7405, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 0.7939403653144836, | |
| "eval_runtime": 30.0181, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.2005169838666916, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.7265, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.1854984015226364, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.689, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.19494923949241638, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.7084, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.7893857955932617, | |
| "eval_runtime": 30.0163, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.19679751992225647, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.6769, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.23057012259960175, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.6615, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 0.7893829345703125, | |
| "eval_runtime": 30.0052, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.22995226085186005, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.677, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.2545529305934906, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.6218, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.2671603262424469, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.6539, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.7934186458587646, | |
| "eval_runtime": 30.0147, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.27313029766082764, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.5984, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.4352659285068512, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.5823, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.30285102128982544, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.5975, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 0.8043038249015808, | |
| "eval_runtime": 30.008, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.32653096318244934, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.5633, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.3424088656902313, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.5581, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.4328867793083191, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.5255, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.8410797119140625, | |
| "eval_runtime": 30.0137, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.7682741284370422, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.4971, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 0.7199717164039612, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.4808, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 0.8795941472053528, | |
| "eval_runtime": 30.0151, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.8249652981758118, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.4617, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.7966895699501038, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.4737, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 0.653789758682251, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.4272, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.9054422378540039, | |
| "eval_runtime": 30.0127, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.181818181818182, | |
| "grad_norm": 0.6517956256866455, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.3816, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 14.545454545454545, | |
| "grad_norm": 0.9160972833633423, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.4123, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "grad_norm": 0.7034734487533569, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.3206, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "eval_loss": 0.94171142578125, | |
| "eval_runtime": 30.012, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 15.272727272727273, | |
| "grad_norm": 0.8707321286201477, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.3492, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 15.636363636363637, | |
| "grad_norm": 0.861784815788269, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.3156, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.818087100982666, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 0.3136, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.9864615797996521, | |
| "eval_runtime": 30.0165, | |
| "eval_samples_per_second": 0.8, | |
| "eval_steps_per_second": 0.8, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "step": 88, | |
| "total_flos": 1.6598131547098317e+17, | |
| "train_loss": 0.6967281076041135, | |
| "train_runtime": 6042.6364, | |
| "train_samples_per_second": 0.728, | |
| "train_steps_per_second": 0.041 | |
| } | |
| ], | |
| "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": 1.6598131547098317e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |