| { | |
| "best_metric": 0.9981787800788879, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-2_120-samples_config-2_full_auto/checkpoint-71", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 110, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.499232679605484, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.5314, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.45545342564582825, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.4846, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.42690765857696533, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.5395, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 1.5119134187698364, | |
| "eval_runtime": 25.0009, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.4263952970504761, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.5219, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.4570830762386322, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.477, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.35217902064323425, | |
| "learning_rate": 4e-05, | |
| "loss": 1.4907, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.417191505432129, | |
| "eval_runtime": 25.0036, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.27743151783943176, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.3811, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.31835851073265076, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.3756, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.22837312519550323, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.3412, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 1.3288416862487793, | |
| "eval_runtime": 24.9916, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.23391933739185333, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.2966, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.25118717551231384, | |
| "learning_rate": 8e-05, | |
| "loss": 1.2484, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.22090202569961548, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.2282, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.2153141498565674, | |
| "eval_runtime": 24.9941, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.22254027426242828, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.1737, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.25803518295288086, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 1.1141, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 1.1135841608047485, | |
| "eval_runtime": 24.9994, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.2485925853252411, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 1.092, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.20472270250320435, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 1.0324, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.17066584527492523, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 1.0131, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.0539630651474, | |
| "eval_runtime": 24.9992, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.14014144241809845, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 1.0045, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.12622663378715515, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.9826, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.12965665757656097, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 1.0044, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 1.0380480289459229, | |
| "eval_runtime": 24.9924, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.131666362285614, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.9948, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.12134098261594772, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.9451, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.11297284811735153, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.9748, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.022283673286438, | |
| "eval_runtime": 25.006, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.13100577890872955, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.9412, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.14722400903701782, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.937, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 1.0142197608947754, | |
| "eval_runtime": 24.9988, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.13433553278446198, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.9592, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.13646681606769562, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.9053, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.14214488863945007, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.9481, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.0053483247756958, | |
| "eval_runtime": 24.9925, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.12895554304122925, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.8967, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.16620439291000366, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.8977, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.13502992689609528, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.9023, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 1.0011383295059204, | |
| "eval_runtime": 25.0027, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.16103532910346985, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.887, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.1596815139055252, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.8935, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.207292839884758, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.8716, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.9987447261810303, | |
| "eval_runtime": 24.9973, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.1768721491098404, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.858, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 0.2096080631017685, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.849, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 0.9981787800788879, | |
| "eval_runtime": 25.0048, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.20918868482112885, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.8323, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.16997645795345306, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.8717, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 0.19361475110054016, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.836, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.0032271146774292, | |
| "eval_runtime": 24.9941, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.181818181818182, | |
| "grad_norm": 0.19343358278274536, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.7833, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 14.545454545454545, | |
| "grad_norm": 0.22505107522010803, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.8369, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "grad_norm": 0.26781898736953735, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.7365, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "eval_loss": 1.017637848854065, | |
| "eval_runtime": 24.9953, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 15.272727272727273, | |
| "grad_norm": 0.2967229187488556, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.7896, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 15.636363636363637, | |
| "grad_norm": 0.34663647413253784, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.746, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.34494930505752563, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 0.7495, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.0259361267089844, | |
| "eval_runtime": 24.995, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.363636363636363, | |
| "grad_norm": 0.2989581823348999, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.7907, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 16.727272727272727, | |
| "grad_norm": 0.5614259839057922, | |
| "learning_rate": 7.967094433018508e-05, | |
| "loss": 0.6882, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 16.90909090909091, | |
| "eval_loss": 1.0410116910934448, | |
| "eval_runtime": 25.0091, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 17.09090909090909, | |
| "grad_norm": 0.46929559111595154, | |
| "learning_rate": 7.85356783842216e-05, | |
| "loss": 0.6714, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 17.454545454545453, | |
| "grad_norm": 0.494697630405426, | |
| "learning_rate": 7.737816117462752e-05, | |
| "loss": 0.6695, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 17.818181818181817, | |
| "grad_norm": 0.4196230471134186, | |
| "learning_rate": 7.619929529850397e-05, | |
| "loss": 0.6301, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 1.070792317390442, | |
| "eval_runtime": 24.9946, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 18.181818181818183, | |
| "grad_norm": 0.43488675355911255, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.671, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 18.545454545454547, | |
| "grad_norm": 0.6978307962417603, | |
| "learning_rate": 7.378121045351378e-05, | |
| "loss": 0.6072, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 18.90909090909091, | |
| "grad_norm": 0.4898715913295746, | |
| "learning_rate": 7.254387703447154e-05, | |
| "loss": 0.6114, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 18.90909090909091, | |
| "eval_loss": 1.1067150831222534, | |
| "eval_runtime": 24.9953, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 19.272727272727273, | |
| "grad_norm": 0.4548497796058655, | |
| "learning_rate": 7.128896457825364e-05, | |
| "loss": 0.5898, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 19.636363636363637, | |
| "grad_norm": 0.4812833368778229, | |
| "learning_rate": 7.001745162784477e-05, | |
| "loss": 0.5549, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.5569318532943726, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 0.591, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 1.133439064025879, | |
| "eval_runtime": 24.9952, | |
| "eval_samples_per_second": 0.96, | |
| "eval_steps_per_second": 0.96, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 110, | |
| "total_flos": 1.7419861205660467e+17, | |
| "train_loss": 0.9589717778292569, | |
| "train_runtime": 6212.9293, | |
| "train_samples_per_second": 0.708, | |
| "train_steps_per_second": 0.04 | |
| } | |
| ], | |
| "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": 1.7419861205660467e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |