| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.991150442477876, | |
| "eval_steps": 500, | |
| "global_step": 252, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.058997050147492625, | |
| "grad_norm": 2.2326040267944336, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 0.9779, | |
| "mean_token_accuracy": 0.7360098913311959, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.11799410029498525, | |
| "grad_norm": 0.895714282989502, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.8643, | |
| "mean_token_accuracy": 0.7512570098042488, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.17699115044247787, | |
| "grad_norm": 0.639182448387146, | |
| "learning_rate": 4.9992225160291885e-05, | |
| "loss": 0.7757, | |
| "mean_token_accuracy": 0.7676257744431496, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2359882005899705, | |
| "grad_norm": 0.476315975189209, | |
| "learning_rate": 4.990481990931694e-05, | |
| "loss": 0.74, | |
| "mean_token_accuracy": 0.774705457687378, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2949852507374631, | |
| "grad_norm": 0.4143844246864319, | |
| "learning_rate": 4.9720669537317554e-05, | |
| "loss": 0.7214, | |
| "mean_token_accuracy": 0.7784379974007607, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.3763030469417572, | |
| "learning_rate": 4.944056921471231e-05, | |
| "loss": 0.7044, | |
| "mean_token_accuracy": 0.7822064161300659, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.41297935103244837, | |
| "grad_norm": 0.34689950942993164, | |
| "learning_rate": 4.9065728428554405e-05, | |
| "loss": 0.6841, | |
| "mean_token_accuracy": 0.7874152660369873, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.471976401179941, | |
| "grad_norm": 0.332253098487854, | |
| "learning_rate": 4.859776575990668e-05, | |
| "loss": 0.6713, | |
| "mean_token_accuracy": 0.7901755809783936, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5309734513274337, | |
| "grad_norm": 0.33200857043266296, | |
| "learning_rate": 4.8038701894728526e-05, | |
| "loss": 0.6553, | |
| "mean_token_accuracy": 0.7945165485143661, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5899705014749262, | |
| "grad_norm": 0.3228710889816284, | |
| "learning_rate": 4.7390950898453896e-05, | |
| "loss": 0.6583, | |
| "mean_token_accuracy": 0.7935768470168114, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6489675516224189, | |
| "grad_norm": 0.34554317593574524, | |
| "learning_rate": 4.665730979193731e-05, | |
| "loss": 0.6575, | |
| "mean_token_accuracy": 0.7934844985604286, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.32089245319366455, | |
| "learning_rate": 4.584094647377942e-05, | |
| "loss": 0.6508, | |
| "mean_token_accuracy": 0.795045793056488, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7669616519174042, | |
| "grad_norm": 0.36928287148475647, | |
| "learning_rate": 4.494538604118408e-05, | |
| "loss": 0.649, | |
| "mean_token_accuracy": 0.795442347228527, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8259587020648967, | |
| "grad_norm": 0.3632085919380188, | |
| "learning_rate": 4.3974495568414175e-05, | |
| "loss": 0.6488, | |
| "mean_token_accuracy": 0.795751303434372, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "grad_norm": 0.3389696776866913, | |
| "learning_rate": 4.293246740857364e-05, | |
| "loss": 0.6398, | |
| "mean_token_accuracy": 0.7981088876724243, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.943952802359882, | |
| "grad_norm": 0.32345691323280334, | |
| "learning_rate": 4.182380109081926e-05, | |
| "loss": 0.6407, | |
| "mean_token_accuracy": 0.7974872574210167, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0117994100294985, | |
| "grad_norm": 0.7706670165061951, | |
| "learning_rate": 4.065328389117115e-05, | |
| "loss": 0.7546, | |
| "mean_token_accuracy": 0.8011921731437125, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.0707964601769913, | |
| "grad_norm": 0.405765563249588, | |
| "learning_rate": 3.9425970160818e-05, | |
| "loss": 0.5884, | |
| "mean_token_accuracy": 0.8105607956647873, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1297935103244838, | |
| "grad_norm": 0.307887464761734, | |
| "learning_rate": 3.814715950117856e-05, | |
| "loss": 0.594, | |
| "mean_token_accuracy": 0.808464927971363, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.1887905604719764, | |
| "grad_norm": 0.3324905037879944, | |
| "learning_rate": 3.682237387996043e-05, | |
| "loss": 0.5868, | |
| "mean_token_accuracy": 0.8108821675181389, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.247787610619469, | |
| "grad_norm": 0.342385470867157, | |
| "learning_rate": 3.545733378703032e-05, | |
| "loss": 0.5885, | |
| "mean_token_accuracy": 0.8102918788790703, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3067846607669615, | |
| "grad_norm": 0.28968796133995056, | |
| "learning_rate": 3.405793353305565e-05, | |
| "loss": 0.5855, | |
| "mean_token_accuracy": 0.8110313430428505, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.3657817109144543, | |
| "grad_norm": 0.372152715921402, | |
| "learning_rate": 3.2630215797579524e-05, | |
| "loss": 0.5919, | |
| "mean_token_accuracy": 0.8090123891830444, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.424778761061947, | |
| "grad_norm": 0.2889751195907593, | |
| "learning_rate": 3.1180345536431366e-05, | |
| "loss": 0.5883, | |
| "mean_token_accuracy": 0.8098931536078453, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.4837758112094395, | |
| "grad_norm": 0.3183657228946686, | |
| "learning_rate": 2.971458336114263e-05, | |
| "loss": 0.5791, | |
| "mean_token_accuracy": 0.8125559493899346, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.5427728613569323, | |
| "grad_norm": 0.31042715907096863, | |
| "learning_rate": 2.8239258505316425e-05, | |
| "loss": 0.5852, | |
| "mean_token_accuracy": 0.8108706936240196, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6017699115044248, | |
| "grad_norm": 0.33204129338264465, | |
| "learning_rate": 2.6760741494683584e-05, | |
| "loss": 0.5746, | |
| "mean_token_accuracy": 0.8140962585806847, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.6607669616519174, | |
| "grad_norm": 0.342929482460022, | |
| "learning_rate": 2.5285416638857372e-05, | |
| "loss": 0.575, | |
| "mean_token_accuracy": 0.8141591981053352, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.71976401179941, | |
| "grad_norm": 0.3200264573097229, | |
| "learning_rate": 2.3819654463568643e-05, | |
| "loss": 0.574, | |
| "mean_token_accuracy": 0.8143440678715705, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.7787610619469025, | |
| "grad_norm": 0.287501722574234, | |
| "learning_rate": 2.2369784202420478e-05, | |
| "loss": 0.5774, | |
| "mean_token_accuracy": 0.8128512471914291, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.8377581120943953, | |
| "grad_norm": 0.2589595317840576, | |
| "learning_rate": 2.0942066466944355e-05, | |
| "loss": 0.5764, | |
| "mean_token_accuracy": 0.8133013695478439, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.896755162241888, | |
| "grad_norm": 0.32855328917503357, | |
| "learning_rate": 1.9542666212969686e-05, | |
| "loss": 0.5767, | |
| "mean_token_accuracy": 0.8132460281252861, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.9557522123893807, | |
| "grad_norm": 0.26567938923835754, | |
| "learning_rate": 1.817762612003957e-05, | |
| "loss": 0.5783, | |
| "mean_token_accuracy": 0.8122945606708527, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.023598820058997, | |
| "grad_norm": 0.3464497923851013, | |
| "learning_rate": 1.685284049882144e-05, | |
| "loss": 0.6753, | |
| "mean_token_accuracy": 0.8177848967110238, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.0825958702064895, | |
| "grad_norm": 0.3261229395866394, | |
| "learning_rate": 1.5574029839182002e-05, | |
| "loss": 0.5317, | |
| "mean_token_accuracy": 0.8253509551286697, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.1415929203539825, | |
| "grad_norm": 0.2763068377971649, | |
| "learning_rate": 1.4346716108828856e-05, | |
| "loss": 0.5282, | |
| "mean_token_accuracy": 0.8263520315289498, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.200589970501475, | |
| "grad_norm": 0.2966022193431854, | |
| "learning_rate": 1.3176198909180748e-05, | |
| "loss": 0.53, | |
| "mean_token_accuracy": 0.8254068896174431, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.2595870206489677, | |
| "grad_norm": 0.2522096633911133, | |
| "learning_rate": 1.2067532591426361e-05, | |
| "loss": 0.5294, | |
| "mean_token_accuracy": 0.8256875559687614, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.3185840707964602, | |
| "grad_norm": 0.23973774909973145, | |
| "learning_rate": 1.1025504431585829e-05, | |
| "loss": 0.5343, | |
| "mean_token_accuracy": 0.8240553349256515, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.377581120943953, | |
| "grad_norm": 0.2457086741924286, | |
| "learning_rate": 1.0054613958815923e-05, | |
| "loss": 0.5311, | |
| "mean_token_accuracy": 0.825106480717659, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.4365781710914454, | |
| "grad_norm": 0.23020589351654053, | |
| "learning_rate": 9.159053526220584e-06, | |
| "loss": 0.5274, | |
| "mean_token_accuracy": 0.8265491768717765, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.495575221238938, | |
| "grad_norm": 0.23255181312561035, | |
| "learning_rate": 8.342690208062693e-06, | |
| "loss": 0.5356, | |
| "mean_token_accuracy": 0.8239412903785706, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.554572271386431, | |
| "grad_norm": 0.22837501764297485, | |
| "learning_rate": 7.609049101546113e-06, | |
| "loss": 0.5346, | |
| "mean_token_accuracy": 0.8239208117127419, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.613569321533923, | |
| "grad_norm": 0.2300056517124176, | |
| "learning_rate": 6.961298105271478e-06, | |
| "loss": 0.531, | |
| "mean_token_accuracy": 0.8248263359069824, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.672566371681416, | |
| "grad_norm": 0.22930695116519928, | |
| "learning_rate": 6.40223424009332e-06, | |
| "loss": 0.533, | |
| "mean_token_accuracy": 0.8248327806591987, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.7315634218289087, | |
| "grad_norm": 0.20990419387817383, | |
| "learning_rate": 5.934271571445601e-06, | |
| "loss": 0.5287, | |
| "mean_token_accuracy": 0.825985960662365, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.7905604719764012, | |
| "grad_norm": 0.22643804550170898, | |
| "learning_rate": 5.559430785287692e-06, | |
| "loss": 0.5322, | |
| "mean_token_accuracy": 0.8249855458736419, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.849557522123894, | |
| "grad_norm": 0.21520549058914185, | |
| "learning_rate": 5.279330462682451e-06, | |
| "loss": 0.5316, | |
| "mean_token_accuracy": 0.8252911075949669, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.9085545722713864, | |
| "grad_norm": 0.2218010574579239, | |
| "learning_rate": 5.095180090683062e-06, | |
| "loss": 0.5322, | |
| "mean_token_accuracy": 0.8245415687561035, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.967551622418879, | |
| "grad_norm": 0.20834492146968842, | |
| "learning_rate": 5.007774839708124e-06, | |
| "loss": 0.5181, | |
| "mean_token_accuracy": 0.8292100727558136, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.991150442477876, | |
| "mean_token_accuracy": 0.8264780156314373, | |
| "step": 252, | |
| "total_flos": 383335494844416.0, | |
| "train_loss": 0.6108051699305338, | |
| "train_runtime": 6955.6722, | |
| "train_samples_per_second": 4.672, | |
| "train_steps_per_second": 0.036 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 252, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 383335494844416.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |