| { | |
| "best_metric": 0.8633480668067932, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-2_120-samples_config-2_full_auto/checkpoint-66", | |
| "epoch": 18.90909090909091, | |
| "eval_steps": 500, | |
| "global_step": 104, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.4822298288345337, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.3344, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.4004879593849182, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.2637, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.4426698386669159, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.3338, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 1.3081936836242676, | |
| "eval_runtime": 9.6571, | |
| "eval_samples_per_second": 2.485, | |
| "eval_steps_per_second": 2.485, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.46916842460632324, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.3096, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.44375067949295044, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.2931, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.4392260015010834, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3005, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.2186788320541382, | |
| "eval_runtime": 9.3955, | |
| "eval_samples_per_second": 2.554, | |
| "eval_steps_per_second": 2.554, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.30896463990211487, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.1793, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.27442148327827454, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.1689, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.24738050997257233, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.1249, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 1.1210649013519287, | |
| "eval_runtime": 9.3887, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.25026676058769226, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.1109, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.2738231122493744, | |
| "learning_rate": 8e-05, | |
| "loss": 1.049, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.23668889701366425, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.0441, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.0411239862442017, | |
| "eval_runtime": 9.3933, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.23431846499443054, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.0031, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.27445587515830994, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.9601, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 0.9654064178466797, | |
| "eval_runtime": 9.3966, | |
| "eval_samples_per_second": 2.554, | |
| "eval_steps_per_second": 2.554, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.2644496262073517, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.93, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.19426663219928741, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.8963, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.16398510336875916, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.8731, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.9164630770683289, | |
| "eval_runtime": 9.3903, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.18445228040218353, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.862, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.14853909611701965, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.8378, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.14185050129890442, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.8553, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 0.8962129950523376, | |
| "eval_runtime": 9.3906, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.13734756410121918, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.8468, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.12671595811843872, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.797, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.11967266350984573, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.8229, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.8815064430236816, | |
| "eval_runtime": 9.384, | |
| "eval_samples_per_second": 2.558, | |
| "eval_steps_per_second": 2.558, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.1320078819990158, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.7905, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.1441613882780075, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.7958, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 0.8724080920219421, | |
| "eval_runtime": 9.3831, | |
| "eval_samples_per_second": 2.558, | |
| "eval_steps_per_second": 2.558, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.13146449625492096, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.8017, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.1331482231616974, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.7598, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 4.113140106201172, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.7945, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.8658233284950256, | |
| "eval_runtime": 9.3888, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.13863505423069, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.7521, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.13574767112731934, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.7508, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.14597807824611664, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.7557, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 0.8633802533149719, | |
| "eval_runtime": 9.4007, | |
| "eval_samples_per_second": 2.553, | |
| "eval_steps_per_second": 2.553, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.1450115144252777, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.7375, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.14700905978679657, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.7465, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.1770864874124527, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.7322, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.8633480668067932, | |
| "eval_runtime": 9.3872, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.16425994038581848, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.7137, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 0.1712324172258377, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.7081, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 0.8663877844810486, | |
| "eval_runtime": 9.3945, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.18763937056064606, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.6958, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.17091812193393707, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.7258, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 0.18959878385066986, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.6938, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.8732501864433289, | |
| "eval_runtime": 9.3874, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.181818181818182, | |
| "grad_norm": 0.19071567058563232, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.6512, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 14.545454545454545, | |
| "grad_norm": 0.2350921630859375, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.6989, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "grad_norm": 0.2526806592941284, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.6136, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "eval_loss": 0.8832157254219055, | |
| "eval_runtime": 9.3971, | |
| "eval_samples_per_second": 2.554, | |
| "eval_steps_per_second": 2.554, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 15.272727272727273, | |
| "grad_norm": 0.22328150272369385, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.6554, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 15.636363636363637, | |
| "grad_norm": 0.2550279498100281, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.6237, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.2746519148349762, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 0.6249, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.8979750275611877, | |
| "eval_runtime": 9.385, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.363636363636363, | |
| "grad_norm": 0.2654627561569214, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.6601, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 16.727272727272727, | |
| "grad_norm": 0.2994433641433716, | |
| "learning_rate": 7.967094433018508e-05, | |
| "loss": 0.573, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 16.90909090909091, | |
| "eval_loss": 0.9280126690864563, | |
| "eval_runtime": 9.3912, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 17.09090909090909, | |
| "grad_norm": 0.3651883602142334, | |
| "learning_rate": 7.85356783842216e-05, | |
| "loss": 0.5659, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 17.454545454545453, | |
| "grad_norm": 0.38829582929611206, | |
| "learning_rate": 7.737816117462752e-05, | |
| "loss": 0.5625, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 17.818181818181817, | |
| "grad_norm": 0.6042073965072632, | |
| "learning_rate": 7.619929529850397e-05, | |
| "loss": 0.5309, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 0.9384562969207764, | |
| "eval_runtime": 9.3905, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 18.181818181818183, | |
| "grad_norm": 0.47595301270484924, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.5669, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 18.545454545454547, | |
| "grad_norm": 0.4337148070335388, | |
| "learning_rate": 7.378121045351378e-05, | |
| "loss": 0.5193, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 18.90909090909091, | |
| "grad_norm": 0.4653347432613373, | |
| "learning_rate": 7.254387703447154e-05, | |
| "loss": 0.5198, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 18.90909090909091, | |
| "eval_loss": 0.9873903393745422, | |
| "eval_runtime": 9.3954, | |
| "eval_samples_per_second": 2.554, | |
| "eval_steps_per_second": 2.554, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 18.90909090909091, | |
| "step": 104, | |
| "total_flos": 5.622222715407565e+16, | |
| "train_loss": 0.8338750004768372, | |
| "train_runtime": 2406.1658, | |
| "train_samples_per_second": 1.829, | |
| "train_steps_per_second": 0.104 | |
| } | |
| ], | |
| "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": 5.622222715407565e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |