| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 6, | |
| "global_step": 63, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.047619047619047616, | |
| "grad_norm": 8.375894546508789, | |
| "learning_rate": 1e-05, | |
| "loss": 8.6884, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.047619047619047616, | |
| "eval_loss": 8.88193130493164, | |
| "eval_runtime": 2.7491, | |
| "eval_samples_per_second": 6.548, | |
| "eval_steps_per_second": 1.091, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.09523809523809523, | |
| "grad_norm": 8.514049530029297, | |
| "learning_rate": 2e-05, | |
| "loss": 8.763, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 8.759590148925781, | |
| "learning_rate": 3e-05, | |
| "loss": 8.8678, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.19047619047619047, | |
| "grad_norm": 9.424858093261719, | |
| "learning_rate": 4e-05, | |
| "loss": 8.7207, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.23809523809523808, | |
| "grad_norm": 9.865653038024902, | |
| "learning_rate": 5e-05, | |
| "loss": 8.5683, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 10.115079879760742, | |
| "learning_rate": 6e-05, | |
| "loss": 8.0217, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "eval_loss": 7.370055198669434, | |
| "eval_runtime": 2.2437, | |
| "eval_samples_per_second": 8.022, | |
| "eval_steps_per_second": 1.337, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 11.35451889038086, | |
| "learning_rate": 7e-05, | |
| "loss": 7.197, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.38095238095238093, | |
| "grad_norm": 12.043075561523438, | |
| "learning_rate": 8e-05, | |
| "loss": 6.5971, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 13.008752822875977, | |
| "learning_rate": 9e-05, | |
| "loss": 5.2664, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.47619047619047616, | |
| "grad_norm": 12.774999618530273, | |
| "learning_rate": 0.0001, | |
| "loss": 3.7546, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5238095238095238, | |
| "grad_norm": 9.925054550170898, | |
| "learning_rate": 9.991218658821608e-05, | |
| "loss": 2.8489, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 10.283626556396484, | |
| "learning_rate": 9.964905480067586e-05, | |
| "loss": 1.647, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "eval_loss": 0.8834102153778076, | |
| "eval_runtime": 2.2436, | |
| "eval_samples_per_second": 8.023, | |
| "eval_steps_per_second": 1.337, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.6190476190476191, | |
| "grad_norm": 9.272812843322754, | |
| "learning_rate": 9.921152889737984e-05, | |
| "loss": 0.7634, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 6.9024434089660645, | |
| "learning_rate": 9.860114570402054e-05, | |
| "loss": 0.2569, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 1.7125144004821777, | |
| "learning_rate": 9.782004921382612e-05, | |
| "loss": 0.034, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 2.1755614280700684, | |
| "learning_rate": 9.687098305670605e-05, | |
| "loss": 0.1234, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.8095238095238095, | |
| "grad_norm": 0.0599493645131588, | |
| "learning_rate": 9.575728086215092e-05, | |
| "loss": 0.0014, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 0.03578714281320572, | |
| "learning_rate": 9.448285454973738e-05, | |
| "loss": 0.0009, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "eval_loss": 0.0005784661043435335, | |
| "eval_runtime": 2.2484, | |
| "eval_samples_per_second": 8.006, | |
| "eval_steps_per_second": 1.334, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.9047619047619048, | |
| "grad_norm": 0.019501160830259323, | |
| "learning_rate": 9.305218058836778e-05, | |
| "loss": 0.0005, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.9523809523809523, | |
| "grad_norm": 0.022740766406059265, | |
| "learning_rate": 9.14702842725101e-05, | |
| "loss": 0.0005, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.259290874004364, | |
| "learning_rate": 8.974272207066767e-05, | |
| "loss": 0.0013, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 1.0476190476190477, | |
| "grad_norm": 0.18333713710308075, | |
| "learning_rate": 8.787556210808101e-05, | |
| "loss": 0.001, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.0952380952380953, | |
| "grad_norm": 0.034940749406814575, | |
| "learning_rate": 8.587536285221656e-05, | |
| "loss": 0.0005, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 0.031384099274873734, | |
| "learning_rate": 8.374915007591053e-05, | |
| "loss": 0.0005, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "eval_loss": 0.0014465778367593884, | |
| "eval_runtime": 2.2674, | |
| "eval_samples_per_second": 7.939, | |
| "eval_steps_per_second": 1.323, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.1904761904761905, | |
| "grad_norm": 0.019181104376912117, | |
| "learning_rate": 8.150439217908556e-05, | |
| "loss": 0.0003, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.2380952380952381, | |
| "grad_norm": 0.02535114623606205, | |
| "learning_rate": 7.91489739557236e-05, | |
| "loss": 0.0003, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.2857142857142856, | |
| "grad_norm": 0.08832518011331558, | |
| "learning_rate": 7.669116889823955e-05, | |
| "loss": 0.0009, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.02784010022878647, | |
| "learning_rate": 7.413961013653726e-05, | |
| "loss": 0.0004, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.380952380952381, | |
| "grad_norm": 0.024539506062865257, | |
| "learning_rate": 7.150326011382604e-05, | |
| "loss": 0.0003, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 0.04319535195827484, | |
| "learning_rate": 6.879137910571191e-05, | |
| "loss": 0.0004, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "eval_loss": 0.0004903983790427446, | |
| "eval_runtime": 2.2497, | |
| "eval_samples_per_second": 8.001, | |
| "eval_steps_per_second": 1.334, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.4761904761904763, | |
| "grad_norm": 0.014017721638083458, | |
| "learning_rate": 6.601349269314188e-05, | |
| "loss": 0.0003, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.5238095238095237, | |
| "grad_norm": 0.0841885432600975, | |
| "learning_rate": 6.317935830345338e-05, | |
| "loss": 0.0006, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.5714285714285714, | |
| "grad_norm": 0.020319445058703423, | |
| "learning_rate": 6.029893093705492e-05, | |
| "loss": 0.0003, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.619047619047619, | |
| "grad_norm": 1.9485701322555542, | |
| "learning_rate": 5.738232820012407e-05, | |
| "loss": 0.0164, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.02285480685532093, | |
| "learning_rate": 5.4439794766146746e-05, | |
| "loss": 0.0003, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "grad_norm": 0.16724388301372528, | |
| "learning_rate": 5.148166639112799e-05, | |
| "loss": 0.001, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.7142857142857144, | |
| "eval_loss": 0.0006431459914892912, | |
| "eval_runtime": 2.2493, | |
| "eval_samples_per_second": 8.003, | |
| "eval_steps_per_second": 1.334, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.7619047619047619, | |
| "grad_norm": 0.047226764261722565, | |
| "learning_rate": 4.851833360887201e-05, | |
| "loss": 0.0004, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.8095238095238095, | |
| "grad_norm": 0.1621876060962677, | |
| "learning_rate": 4.5560205233853266e-05, | |
| "loss": 0.0012, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.8571428571428572, | |
| "grad_norm": 0.0308663472533226, | |
| "learning_rate": 4.2617671799875944e-05, | |
| "loss": 0.0003, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.9047619047619047, | |
| "grad_norm": 0.011845704168081284, | |
| "learning_rate": 3.970106906294509e-05, | |
| "loss": 0.0002, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.9523809523809523, | |
| "grad_norm": 0.003460651496425271, | |
| "learning_rate": 3.682064169654663e-05, | |
| "loss": 0.0001, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.012540038675069809, | |
| "learning_rate": 3.3986507306858125e-05, | |
| "loss": 0.0002, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.0001648629695409909, | |
| "eval_runtime": 2.2501, | |
| "eval_samples_per_second": 7.999, | |
| "eval_steps_per_second": 1.333, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.0476190476190474, | |
| "grad_norm": 0.0032971783075481653, | |
| "learning_rate": 3.12086208942881e-05, | |
| "loss": 0.0001, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 2.0952380952380953, | |
| "grad_norm": 0.0028387894853949547, | |
| "learning_rate": 2.8496739886173995e-05, | |
| "loss": 0.0001, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 0.004313192795962095, | |
| "learning_rate": 2.5860389863462765e-05, | |
| "loss": 0.0001, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.1904761904761907, | |
| "grad_norm": 0.002875348087400198, | |
| "learning_rate": 2.3308831101760486e-05, | |
| "loss": 0.0001, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 2.238095238095238, | |
| "grad_norm": 0.002736931899562478, | |
| "learning_rate": 2.0851026044276406e-05, | |
| "loss": 0.0001, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 0.002644361462444067, | |
| "learning_rate": 1.849560782091445e-05, | |
| "loss": 0.0001, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "eval_loss": 0.00020012857567053288, | |
| "eval_runtime": 2.2588, | |
| "eval_samples_per_second": 7.969, | |
| "eval_steps_per_second": 1.328, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.002743582706898451, | |
| "learning_rate": 1.6250849924089484e-05, | |
| "loss": 0.0001, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 2.380952380952381, | |
| "grad_norm": 0.002709955209866166, | |
| "learning_rate": 1.4124637147783432e-05, | |
| "loss": 0.0001, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.4285714285714284, | |
| "grad_norm": 0.0027039337437599897, | |
| "learning_rate": 1.2124437891918993e-05, | |
| "loss": 0.0001, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.4761904761904763, | |
| "grad_norm": 0.002979928394779563, | |
| "learning_rate": 1.0257277929332332e-05, | |
| "loss": 0.0001, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.5238095238095237, | |
| "grad_norm": 0.002763009862974286, | |
| "learning_rate": 8.529715727489912e-06, | |
| "loss": 0.0001, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "grad_norm": 0.0026082261465489864, | |
| "learning_rate": 6.947819411632223e-06, | |
| "loss": 0.0001, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.571428571428571, | |
| "eval_loss": 0.00020980079716537148, | |
| "eval_runtime": 2.2421, | |
| "eval_samples_per_second": 8.028, | |
| "eval_steps_per_second": 1.338, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.619047619047619, | |
| "grad_norm": 0.0024540037848055363, | |
| "learning_rate": 5.51714545026264e-06, | |
| "loss": 0.0001, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.002579254098236561, | |
| "learning_rate": 4.242719137849077e-06, | |
| "loss": 0.0001, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.7142857142857144, | |
| "grad_norm": 0.0024759070947766304, | |
| "learning_rate": 3.1290169432939553e-06, | |
| "loss": 0.0001, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.761904761904762, | |
| "grad_norm": 0.002558476058766246, | |
| "learning_rate": 2.179950786173879e-06, | |
| "loss": 0.0001, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.8095238095238093, | |
| "grad_norm": 0.0033246162347495556, | |
| "learning_rate": 1.3988542959794627e-06, | |
| "loss": 0.0001, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 0.002528395503759384, | |
| "learning_rate": 7.884711026201585e-07, | |
| "loss": 0.0001, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "eval_loss": 0.0002239210152765736, | |
| "eval_runtime": 2.2503, | |
| "eval_samples_per_second": 7.999, | |
| "eval_steps_per_second": 1.333, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.9047619047619047, | |
| "grad_norm": 0.00311569613404572, | |
| "learning_rate": 3.5094519932415417e-07, | |
| "loss": 0.0001, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.9523809523809526, | |
| "grad_norm": 0.0025822578463703394, | |
| "learning_rate": 8.781341178393244e-08, | |
| "loss": 0.0001, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.02102760784327984, | |
| "learning_rate": 0.0, | |
| "loss": 0.0002, | |
| "step": 63 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 63, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.304496956427469e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |