| { | |
| "best_metric": 0.8825053995680345, | |
| "best_model_checkpoint": "./sentiment_model/checkpoint-3129", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 3576, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11185682326621924, | |
| "grad_norm": 18.68804168701172, | |
| "learning_rate": 5.3691275167785235e-06, | |
| "loss": 4.4686, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22371364653243847, | |
| "grad_norm": 33.30501937866211, | |
| "learning_rate": 1.0738255033557047e-05, | |
| "loss": 4.0139, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.33557046979865773, | |
| "grad_norm": 48.02063751220703, | |
| "learning_rate": 1.633109619686801e-05, | |
| "loss": 2.9933, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.44742729306487694, | |
| "grad_norm": 35.39257049560547, | |
| "learning_rate": 2.192393736017897e-05, | |
| "loss": 2.5147, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5592841163310962, | |
| "grad_norm": 39.318363189697266, | |
| "learning_rate": 2.7404921700223713e-05, | |
| "loss": 2.4221, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6711409395973155, | |
| "grad_norm": 50.962833404541016, | |
| "learning_rate": 3.2997762863534674e-05, | |
| "loss": 2.3124, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7829977628635347, | |
| "grad_norm": 37.32981491088867, | |
| "learning_rate": 3.859060402684564e-05, | |
| "loss": 2.4092, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8948545861297539, | |
| "grad_norm": 32.42803955078125, | |
| "learning_rate": 4.4183445190156604e-05, | |
| "loss": 2.4183, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7844492440604751, | |
| "eval_loss": 0.5040740370750427, | |
| "eval_runtime": 19.6714, | |
| "eval_samples_per_second": 117.683, | |
| "eval_steps_per_second": 3.711, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.0067114093959733, | |
| "grad_norm": 20.918420791625977, | |
| "learning_rate": 4.977628635346757e-05, | |
| "loss": 2.3532, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1185682326621924, | |
| "grad_norm": 28.06084442138672, | |
| "learning_rate": 4.94034302759135e-05, | |
| "loss": 2.1244, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2304250559284116, | |
| "grad_norm": 40.768096923828125, | |
| "learning_rate": 4.8782003479990054e-05, | |
| "loss": 2.3004, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.342281879194631, | |
| "grad_norm": 19.545913696289062, | |
| "learning_rate": 4.816057668406662e-05, | |
| "loss": 2.2399, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.45413870246085, | |
| "grad_norm": 69.29216766357422, | |
| "learning_rate": 4.753914988814318e-05, | |
| "loss": 2.1458, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.5659955257270695, | |
| "grad_norm": 29.00174903869629, | |
| "learning_rate": 4.691772309221974e-05, | |
| "loss": 2.0778, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.6778523489932886, | |
| "grad_norm": 28.77162742614746, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 2.1317, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.7897091722595078, | |
| "grad_norm": 36.04485321044922, | |
| "learning_rate": 4.5674869500372856e-05, | |
| "loss": 2.1345, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.901565995525727, | |
| "grad_norm": 51.31829071044922, | |
| "learning_rate": 4.505344270444942e-05, | |
| "loss": 2.2048, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8138228941684665, | |
| "eval_loss": 0.43246370553970337, | |
| "eval_runtime": 19.9941, | |
| "eval_samples_per_second": 115.784, | |
| "eval_steps_per_second": 3.651, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 2.0134228187919465, | |
| "grad_norm": 19.70203971862793, | |
| "learning_rate": 4.443201590852597e-05, | |
| "loss": 2.0482, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.1252796420581657, | |
| "grad_norm": 44.575721740722656, | |
| "learning_rate": 4.381058911260254e-05, | |
| "loss": 1.8304, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.237136465324385, | |
| "grad_norm": 48.66209411621094, | |
| "learning_rate": 4.31891623166791e-05, | |
| "loss": 1.7784, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.348993288590604, | |
| "grad_norm": 19.984678268432617, | |
| "learning_rate": 4.256773552075566e-05, | |
| "loss": 1.8457, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.460850111856823, | |
| "grad_norm": 64.9332275390625, | |
| "learning_rate": 4.194630872483222e-05, | |
| "loss": 1.7747, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.5727069351230423, | |
| "grad_norm": 20.591524124145508, | |
| "learning_rate": 4.1324881928908774e-05, | |
| "loss": 1.9082, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.684563758389262, | |
| "grad_norm": 24.512027740478516, | |
| "learning_rate": 4.0703455132985336e-05, | |
| "loss": 1.8399, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.796420581655481, | |
| "grad_norm": 18.300504684448242, | |
| "learning_rate": 4.00820283370619e-05, | |
| "loss": 1.7937, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.9082774049217, | |
| "grad_norm": 40.87727737426758, | |
| "learning_rate": 3.946060154113845e-05, | |
| "loss": 1.8174, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8535637149028078, | |
| "eval_loss": 0.37263187766075134, | |
| "eval_runtime": 19.6701, | |
| "eval_samples_per_second": 117.691, | |
| "eval_steps_per_second": 3.711, | |
| "step": 1341 | |
| }, | |
| { | |
| "epoch": 3.0201342281879193, | |
| "grad_norm": 29.54434585571289, | |
| "learning_rate": 3.883917474521502e-05, | |
| "loss": 1.8262, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.131991051454139, | |
| "grad_norm": 19.222951889038086, | |
| "learning_rate": 3.8217747949291576e-05, | |
| "loss": 1.5486, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.243847874720358, | |
| "grad_norm": 36.32904052734375, | |
| "learning_rate": 3.759632115336814e-05, | |
| "loss": 1.5273, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.3557046979865772, | |
| "grad_norm": 27.05373764038086, | |
| "learning_rate": 3.697489435744469e-05, | |
| "loss": 1.5796, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.4675615212527964, | |
| "grad_norm": 50.08060836791992, | |
| "learning_rate": 3.6353467561521254e-05, | |
| "loss": 1.5104, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.5794183445190155, | |
| "grad_norm": 29.98794174194336, | |
| "learning_rate": 3.5732040765597816e-05, | |
| "loss": 1.6298, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.6912751677852347, | |
| "grad_norm": 33.08045196533203, | |
| "learning_rate": 3.511061396967437e-05, | |
| "loss": 1.4769, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.8031319910514543, | |
| "grad_norm": 30.475862503051758, | |
| "learning_rate": 3.448918717375094e-05, | |
| "loss": 1.547, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.9149888143176734, | |
| "grad_norm": 29.52466583251953, | |
| "learning_rate": 3.3867760377827495e-05, | |
| "loss": 1.4918, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8501079913606912, | |
| "eval_loss": 0.38409608602523804, | |
| "eval_runtime": 19.6128, | |
| "eval_samples_per_second": 118.035, | |
| "eval_steps_per_second": 3.722, | |
| "step": 1788 | |
| }, | |
| { | |
| "epoch": 4.026845637583893, | |
| "grad_norm": 27.849456787109375, | |
| "learning_rate": 3.3246333581904056e-05, | |
| "loss": 1.4653, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.138702460850112, | |
| "grad_norm": 39.9719123840332, | |
| "learning_rate": 3.262490678598061e-05, | |
| "loss": 1.2247, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.250559284116331, | |
| "grad_norm": 43.72661590576172, | |
| "learning_rate": 3.200347999005717e-05, | |
| "loss": 1.2065, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.3624161073825505, | |
| "grad_norm": 29.79233169555664, | |
| "learning_rate": 3.1382053194133735e-05, | |
| "loss": 1.2585, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.47427293064877, | |
| "grad_norm": 30.871849060058594, | |
| "learning_rate": 3.076062639821029e-05, | |
| "loss": 1.2602, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.586129753914989, | |
| "grad_norm": 30.835567474365234, | |
| "learning_rate": 3.0139199602286848e-05, | |
| "loss": 1.2781, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.697986577181208, | |
| "grad_norm": 37.51513671875, | |
| "learning_rate": 2.9517772806363413e-05, | |
| "loss": 1.228, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.809843400447427, | |
| "grad_norm": 35.483665466308594, | |
| "learning_rate": 2.889634601043997e-05, | |
| "loss": 1.2876, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 4.921700223713646, | |
| "grad_norm": 21.087987899780273, | |
| "learning_rate": 2.8274919214516533e-05, | |
| "loss": 1.2457, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8660907127429806, | |
| "eval_loss": 0.3854082524776459, | |
| "eval_runtime": 19.9697, | |
| "eval_samples_per_second": 115.926, | |
| "eval_steps_per_second": 3.656, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 5.033557046979865, | |
| "grad_norm": 32.84519577026367, | |
| "learning_rate": 2.765349241859309e-05, | |
| "loss": 1.1247, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.145413870246085, | |
| "grad_norm": 42.308837890625, | |
| "learning_rate": 2.703206562266965e-05, | |
| "loss": 0.9676, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.257270693512305, | |
| "grad_norm": 32.68018341064453, | |
| "learning_rate": 2.6410638826746208e-05, | |
| "loss": 0.9049, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 5.369127516778524, | |
| "grad_norm": 40.90736770629883, | |
| "learning_rate": 2.5789212030822766e-05, | |
| "loss": 0.8973, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.480984340044743, | |
| "grad_norm": 42.41788864135742, | |
| "learning_rate": 2.516778523489933e-05, | |
| "loss": 0.972, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 5.592841163310962, | |
| "grad_norm": 61.866180419921875, | |
| "learning_rate": 2.454635843897589e-05, | |
| "loss": 0.9687, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.704697986577181, | |
| "grad_norm": 54.289512634277344, | |
| "learning_rate": 2.392493164305245e-05, | |
| "loss": 0.986, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 5.8165548098434, | |
| "grad_norm": 32.10563659667969, | |
| "learning_rate": 2.330350484712901e-05, | |
| "loss": 0.981, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.9284116331096195, | |
| "grad_norm": 34.02315139770508, | |
| "learning_rate": 2.2682078051205568e-05, | |
| "loss": 0.9679, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8678185745140389, | |
| "eval_loss": 0.3706440329551697, | |
| "eval_runtime": 19.6479, | |
| "eval_samples_per_second": 117.824, | |
| "eval_steps_per_second": 3.715, | |
| "step": 2682 | |
| }, | |
| { | |
| "epoch": 6.040268456375839, | |
| "grad_norm": 36.86701965332031, | |
| "learning_rate": 2.2060651255282127e-05, | |
| "loss": 0.9318, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.152125279642058, | |
| "grad_norm": 42.711204528808594, | |
| "learning_rate": 2.145165299527716e-05, | |
| "loss": 0.703, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 6.263982102908278, | |
| "grad_norm": 38.12577819824219, | |
| "learning_rate": 2.0830226199353717e-05, | |
| "loss": 0.764, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 6.375838926174497, | |
| "grad_norm": 32.06464767456055, | |
| "learning_rate": 2.0208799403430276e-05, | |
| "loss": 0.739, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 6.487695749440716, | |
| "grad_norm": 36.63414001464844, | |
| "learning_rate": 1.9587372607506838e-05, | |
| "loss": 0.7173, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 6.599552572706935, | |
| "grad_norm": 61.28206253051758, | |
| "learning_rate": 1.8965945811583396e-05, | |
| "loss": 0.7164, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 6.7114093959731544, | |
| "grad_norm": 27.83030891418457, | |
| "learning_rate": 1.8344519015659954e-05, | |
| "loss": 0.7629, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.823266219239374, | |
| "grad_norm": 46.308189392089844, | |
| "learning_rate": 1.7723092219736516e-05, | |
| "loss": 0.7298, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 6.935123042505593, | |
| "grad_norm": 63.58484649658203, | |
| "learning_rate": 1.7101665423813078e-05, | |
| "loss": 0.6876, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8825053995680345, | |
| "eval_loss": 0.3750421702861786, | |
| "eval_runtime": 19.8931, | |
| "eval_samples_per_second": 116.372, | |
| "eval_steps_per_second": 3.67, | |
| "step": 3129 | |
| }, | |
| { | |
| "epoch": 7.046979865771812, | |
| "grad_norm": 44.923614501953125, | |
| "learning_rate": 1.6480238627889636e-05, | |
| "loss": 0.6694, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 7.158836689038031, | |
| "grad_norm": 55.236083984375, | |
| "learning_rate": 1.5858811831966194e-05, | |
| "loss": 0.5705, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 7.27069351230425, | |
| "grad_norm": 57.69952392578125, | |
| "learning_rate": 1.5237385036042756e-05, | |
| "loss": 0.5369, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 7.382550335570469, | |
| "grad_norm": 48.969078063964844, | |
| "learning_rate": 1.4615958240119314e-05, | |
| "loss": 0.5266, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 7.494407158836689, | |
| "grad_norm": 30.38613510131836, | |
| "learning_rate": 1.3994531444195874e-05, | |
| "loss": 0.6426, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 7.6062639821029085, | |
| "grad_norm": 32.06161880493164, | |
| "learning_rate": 1.3385533184190904e-05, | |
| "loss": 0.5366, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 7.718120805369128, | |
| "grad_norm": 41.513572692871094, | |
| "learning_rate": 1.2764106388267464e-05, | |
| "loss": 0.536, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 7.829977628635347, | |
| "grad_norm": 64.65313720703125, | |
| "learning_rate": 1.2142679592344022e-05, | |
| "loss": 0.5542, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 7.941834451901566, | |
| "grad_norm": 59.38785934448242, | |
| "learning_rate": 1.1521252796420582e-05, | |
| "loss": 0.5386, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8825053995680345, | |
| "eval_loss": 0.42663174867630005, | |
| "eval_runtime": 19.6465, | |
| "eval_samples_per_second": 117.833, | |
| "eval_steps_per_second": 3.716, | |
| "step": 3576 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 4470, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.013346575910502e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |