| { | |
| "best_metric": 0.3271898627281189, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-3_120-samples_config-2_auto/checkpoint-55", | |
| "epoch": 16.90909090909091, | |
| "eval_steps": 500, | |
| "global_step": 93, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 3.115290641784668, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.6399, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 2.537904977798462, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.3726, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 2.339669942855835, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.2357, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 2.132227659225464, | |
| "eval_runtime": 21.1364, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 2.310912847518921, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.1577, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 2.905609369277954, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.0476, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 2.008171558380127, | |
| "learning_rate": 4e-05, | |
| "loss": 1.6899, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.3004776239395142, | |
| "eval_runtime": 21.1466, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 2.177971839904785, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.4435, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 1.5493510961532593, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.0639, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 1.4748144149780273, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.6881, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 0.618924081325531, | |
| "eval_runtime": 21.11, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.7245057821273804, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.4158, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.8617022633552551, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5387, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.071713447570801, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.5601, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.44129082560539246, | |
| "eval_runtime": 21.1075, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.9604631662368774, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.4333, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 3.211521863937378, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.3305, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 0.38999900221824646, | |
| "eval_runtime": 21.1088, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.6839059591293335, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.3883, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 1.7793887853622437, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.3067, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.3359837234020233, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.3093, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.35253313183784485, | |
| "eval_runtime": 21.1043, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.19986188411712646, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.3011, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.32240989804267883, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.4026, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.24396806955337524, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.2941, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 0.3439880311489105, | |
| "eval_runtime": 21.1424, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.2701691687107086, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.3205, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 1.857699990272522, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.1969, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.3802018463611603, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.2705, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.33574435114860535, | |
| "eval_runtime": 21.1391, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.35321760177612305, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.2664, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.3299897611141205, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.255, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 0.3277812898159027, | |
| "eval_runtime": 21.2126, | |
| "eval_samples_per_second": 1.131, | |
| "eval_steps_per_second": 1.131, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.27011287212371826, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.2806, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.4219489097595215, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.2605, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.21679142117500305, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.1954, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.3271898627281189, | |
| "eval_runtime": 21.1428, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.2368810772895813, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.1902, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.2612651586532593, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.232, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.23682187497615814, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.1796, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 0.33908501267433167, | |
| "eval_runtime": 21.1408, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.34110820293426514, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.2088, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.259554386138916, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.1467, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.4463067054748535, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.1892, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.36511746048927307, | |
| "eval_runtime": 21.3802, | |
| "eval_samples_per_second": 1.123, | |
| "eval_steps_per_second": 1.123, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.24988265335559845, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.1023, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 0.4166051745414734, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.1795, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 0.3963310420513153, | |
| "eval_runtime": 21.3801, | |
| "eval_samples_per_second": 1.123, | |
| "eval_steps_per_second": 1.123, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.39009249210357666, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.1345, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.3693424165248871, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.0825, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 0.4332319498062134, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.0982, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.4423997402191162, | |
| "eval_runtime": 21.1413, | |
| "eval_samples_per_second": 1.135, | |
| "eval_steps_per_second": 1.135, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.181818181818182, | |
| "grad_norm": 0.36997315287590027, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.1017, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 14.545454545454545, | |
| "grad_norm": 0.44006550312042236, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.069, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "grad_norm": 0.5451624989509583, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.0568, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "eval_loss": 0.5230135321617126, | |
| "eval_runtime": 21.1035, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 15.272727272727273, | |
| "grad_norm": 0.24595630168914795, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.0372, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 15.636363636363637, | |
| "grad_norm": 0.979365348815918, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.026, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.5628170967102051, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 0.046, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.6078879237174988, | |
| "eval_runtime": 21.1325, | |
| "eval_samples_per_second": 1.136, | |
| "eval_steps_per_second": 1.136, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.363636363636363, | |
| "grad_norm": 0.947401762008667, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.019, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 16.727272727272727, | |
| "grad_norm": 0.7989400625228882, | |
| "learning_rate": 7.967094433018508e-05, | |
| "loss": 0.0156, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 16.90909090909091, | |
| "eval_loss": 0.6625682711601257, | |
| "eval_runtime": 21.1062, | |
| "eval_samples_per_second": 1.137, | |
| "eval_steps_per_second": 1.137, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 16.90909090909091, | |
| "step": 93, | |
| "total_flos": 1.1786725109216051e+17, | |
| "train_loss": 0.4795177485994113, | |
| "train_runtime": 4346.2988, | |
| "train_samples_per_second": 1.012, | |
| "train_steps_per_second": 0.058 | |
| } | |
| ], | |
| "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": 1.1786725109216051e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |