| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.300284437220642, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.016253555465258026, | |
| "grad_norm": 2.077186346054077, | |
| "learning_rate": 4.9998691031433496e-05, | |
| "loss": 2.8778, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03250711093051605, | |
| "grad_norm": 2.073504686355591, | |
| "learning_rate": 4.999476426280588e-05, | |
| "loss": 2.6432, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.048760666395774074, | |
| "grad_norm": 2.1281931400299072, | |
| "learning_rate": 4.998822010531848e-05, | |
| "loss": 2.3167, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0650142218610321, | |
| "grad_norm": 1.090219497680664, | |
| "learning_rate": 4.997905924425903e-05, | |
| "loss": 2.1533, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08126777732629012, | |
| "grad_norm": 1.007896900177002, | |
| "learning_rate": 4.996728263892985e-05, | |
| "loss": 2.0267, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09752133279154815, | |
| "grad_norm": 0.9982665777206421, | |
| "learning_rate": 4.995289152254744e-05, | |
| "loss": 1.9352, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11377488825680618, | |
| "grad_norm": 0.8844298720359802, | |
| "learning_rate": 4.9935887402113315e-05, | |
| "loss": 1.9486, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1300284437220642, | |
| "grad_norm": 0.9337536692619324, | |
| "learning_rate": 4.991627205825621e-05, | |
| "loss": 1.9228, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14628199918732224, | |
| "grad_norm": 0.9377800822257996, | |
| "learning_rate": 4.9894047545045605e-05, | |
| "loss": 1.835, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.16253555465258024, | |
| "grad_norm": 0.8525241017341614, | |
| "learning_rate": 4.986921618977664e-05, | |
| "loss": 1.8157, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17878911011783827, | |
| "grad_norm": 0.872871458530426, | |
| "learning_rate": 4.984178059272638e-05, | |
| "loss": 1.8811, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1950426655830963, | |
| "grad_norm": 0.9125804305076599, | |
| "learning_rate": 4.981174362688158e-05, | |
| "loss": 1.8242, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.21129622104835433, | |
| "grad_norm": 0.7474733591079712, | |
| "learning_rate": 4.977910843763777e-05, | |
| "loss": 1.7808, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22754977651361236, | |
| "grad_norm": 1.0248199701309204, | |
| "learning_rate": 4.974387844246987e-05, | |
| "loss": 1.8512, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.24380333197887039, | |
| "grad_norm": 0.9032835960388184, | |
| "learning_rate": 4.970605733057441e-05, | |
| "loss": 1.8172, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2600568874441284, | |
| "grad_norm": 0.8550340533256531, | |
| "learning_rate": 4.9665649062483115e-05, | |
| "loss": 1.7418, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.27631044290938644, | |
| "grad_norm": 0.9429016709327698, | |
| "learning_rate": 4.96226578696482e-05, | |
| "loss": 1.8265, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2925639983746445, | |
| "grad_norm": 0.975885272026062, | |
| "learning_rate": 4.957708825399927e-05, | |
| "loss": 1.7943, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3088175538399025, | |
| "grad_norm": 0.827629804611206, | |
| "learning_rate": 4.9528944987471884e-05, | |
| "loss": 1.8067, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.3250711093051605, | |
| "grad_norm": 1.0019093751907349, | |
| "learning_rate": 4.9478233111507856e-05, | |
| "loss": 1.7874, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3413246647704185, | |
| "grad_norm": 1.0811312198638916, | |
| "learning_rate": 4.9424957936527295e-05, | |
| "loss": 1.7395, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.35757822023567654, | |
| "grad_norm": 0.9461565613746643, | |
| "learning_rate": 4.936912504137257e-05, | |
| "loss": 1.7833, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.37383177570093457, | |
| "grad_norm": 1.0386009216308594, | |
| "learning_rate": 4.9310740272724055e-05, | |
| "loss": 1.7569, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3900853311661926, | |
| "grad_norm": 0.9916568994522095, | |
| "learning_rate": 4.924980974448791e-05, | |
| "loss": 1.7584, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4063388866314506, | |
| "grad_norm": 1.240400791168213, | |
| "learning_rate": 4.918633983715581e-05, | |
| "loss": 1.7438, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.42259244209670865, | |
| "grad_norm": 1.1441287994384766, | |
| "learning_rate": 4.912033719713687e-05, | |
| "loss": 1.7046, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4388459975619667, | |
| "grad_norm": 1.1687984466552734, | |
| "learning_rate": 4.905180873606157e-05, | |
| "loss": 1.738, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.4550995530272247, | |
| "grad_norm": 0.9126319885253906, | |
| "learning_rate": 4.8980761630058014e-05, | |
| "loss": 1.7738, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.47135310849248274, | |
| "grad_norm": 1.172781229019165, | |
| "learning_rate": 4.8907203319000495e-05, | |
| "loss": 1.6642, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.48760666395774077, | |
| "grad_norm": 1.1217776536941528, | |
| "learning_rate": 4.883114150573037e-05, | |
| "loss": 1.7359, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5038602194229987, | |
| "grad_norm": 1.1024829149246216, | |
| "learning_rate": 4.8752584155249444e-05, | |
| "loss": 1.6707, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5201137748882568, | |
| "grad_norm": 1.042017936706543, | |
| "learning_rate": 4.8671539493885934e-05, | |
| "loss": 1.709, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5363673303535148, | |
| "grad_norm": 1.1834771633148193, | |
| "learning_rate": 4.8588016008432945e-05, | |
| "loss": 1.7282, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5526208858187729, | |
| "grad_norm": 1.0311099290847778, | |
| "learning_rate": 4.850202244525987e-05, | |
| "loss": 1.774, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5688744412840309, | |
| "grad_norm": 1.2995566129684448, | |
| "learning_rate": 4.8413567809396376e-05, | |
| "loss": 1.714, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.585127996749289, | |
| "grad_norm": 1.23491370677948, | |
| "learning_rate": 4.8322661363589507e-05, | |
| "loss": 1.7717, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6013815522145469, | |
| "grad_norm": 1.1809489727020264, | |
| "learning_rate": 4.822931262733367e-05, | |
| "loss": 1.722, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.617635107679805, | |
| "grad_norm": 0.988572359085083, | |
| "learning_rate": 4.813353137587377e-05, | |
| "loss": 1.7007, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.633888663145063, | |
| "grad_norm": 1.0803310871124268, | |
| "learning_rate": 4.803532763918162e-05, | |
| "loss": 1.6824, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.650142218610321, | |
| "grad_norm": 1.02713143825531, | |
| "learning_rate": 4.793471170090555e-05, | |
| "loss": 1.6453, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.666395774075579, | |
| "grad_norm": 1.2023831605911255, | |
| "learning_rate": 4.783169409729362e-05, | |
| "loss": 1.6974, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.682649329540837, | |
| "grad_norm": 1.0009326934814453, | |
| "learning_rate": 4.772628561609022e-05, | |
| "loss": 1.7094, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6989028850060951, | |
| "grad_norm": 1.269020438194275, | |
| "learning_rate": 4.761849729540643e-05, | |
| "loss": 1.6832, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7151564404713531, | |
| "grad_norm": 1.3129128217697144, | |
| "learning_rate": 4.7508340422564134e-05, | |
| "loss": 1.6847, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7314099959366112, | |
| "grad_norm": 1.0837377309799194, | |
| "learning_rate": 4.7395826532914054e-05, | |
| "loss": 1.7215, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7476635514018691, | |
| "grad_norm": 1.1211832761764526, | |
| "learning_rate": 4.728096740862778e-05, | |
| "loss": 1.7658, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7639171068671272, | |
| "grad_norm": 1.1422042846679688, | |
| "learning_rate": 4.716377507746397e-05, | |
| "loss": 1.7098, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7801706623323852, | |
| "grad_norm": 1.2925424575805664, | |
| "learning_rate": 4.704426181150884e-05, | |
| "loss": 1.7504, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7964242177976433, | |
| "grad_norm": 1.282771348953247, | |
| "learning_rate": 4.6922440125891064e-05, | |
| "loss": 1.72, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8126777732629012, | |
| "grad_norm": 1.2494529485702515, | |
| "learning_rate": 4.6798322777471216e-05, | |
| "loss": 1.6581, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8289313287281593, | |
| "grad_norm": 1.2736073732376099, | |
| "learning_rate": 4.667192276350591e-05, | |
| "loss": 1.6825, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8451848841934173, | |
| "grad_norm": 1.1477563381195068, | |
| "learning_rate": 4.654325332028676e-05, | |
| "loss": 1.6889, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8614384396586753, | |
| "grad_norm": 1.2415803670883179, | |
| "learning_rate": 4.641232792175428e-05, | |
| "loss": 1.6586, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8776919951239334, | |
| "grad_norm": 1.3159024715423584, | |
| "learning_rate": 4.6279160278086994e-05, | |
| "loss": 1.7102, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8939455505891913, | |
| "grad_norm": 1.2742581367492676, | |
| "learning_rate": 4.614376433426565e-05, | |
| "loss": 1.7476, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9101991060544494, | |
| "grad_norm": 1.34221351146698, | |
| "learning_rate": 4.6006154268613014e-05, | |
| "loss": 1.7497, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9264526615197074, | |
| "grad_norm": 1.1669361591339111, | |
| "learning_rate": 4.586634449130911e-05, | |
| "loss": 1.6693, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9427062169849655, | |
| "grad_norm": 1.2765443325042725, | |
| "learning_rate": 4.572434964288226e-05, | |
| "loss": 1.5862, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9589597724502235, | |
| "grad_norm": 1.2429810762405396, | |
| "learning_rate": 4.55801845926759e-05, | |
| "loss": 1.6875, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9752133279154815, | |
| "grad_norm": 1.2103233337402344, | |
| "learning_rate": 4.543386443729157e-05, | |
| "loss": 1.7294, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9914668833807395, | |
| "grad_norm": 1.3512628078460693, | |
| "learning_rate": 4.528540449900798e-05, | |
| "loss": 1.709, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.0077204388459975, | |
| "grad_norm": 1.2039848566055298, | |
| "learning_rate": 4.513482032417656e-05, | |
| "loss": 1.6632, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.0239739943112556, | |
| "grad_norm": 1.3805475234985352, | |
| "learning_rate": 4.498212768159341e-05, | |
| "loss": 1.695, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.0402275497765137, | |
| "grad_norm": 1.3265151977539062, | |
| "learning_rate": 4.48273425608481e-05, | |
| "loss": 1.6653, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0564811052417715, | |
| "grad_norm": 1.015641212463379, | |
| "learning_rate": 4.4670481170649214e-05, | |
| "loss": 1.652, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0727346607070296, | |
| "grad_norm": 1.2359689474105835, | |
| "learning_rate": 4.451155993712711e-05, | |
| "loss": 1.6457, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0889882161722877, | |
| "grad_norm": 1.3188092708587646, | |
| "learning_rate": 4.435059550211371e-05, | |
| "loss": 1.6512, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.1052417716375458, | |
| "grad_norm": 1.323879361152649, | |
| "learning_rate": 4.4187604721399877e-05, | |
| "loss": 1.6338, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.1214953271028036, | |
| "grad_norm": 1.263047218322754, | |
| "learning_rate": 4.40226046629703e-05, | |
| "loss": 1.6508, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.1377488825680617, | |
| "grad_norm": 1.265649676322937, | |
| "learning_rate": 4.3855612605216174e-05, | |
| "loss": 1.6447, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1540024380333198, | |
| "grad_norm": 1.2379087209701538, | |
| "learning_rate": 4.3686646035125855e-05, | |
| "loss": 1.6522, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.170255993498578, | |
| "grad_norm": 1.200422763824463, | |
| "learning_rate": 4.351572264645366e-05, | |
| "loss": 1.6935, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.1865095489638358, | |
| "grad_norm": 1.3982837200164795, | |
| "learning_rate": 4.334286033786704e-05, | |
| "loss": 1.6792, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.2027631044290938, | |
| "grad_norm": 1.3108696937561035, | |
| "learning_rate": 4.3168077211072263e-05, | |
| "loss": 1.6498, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.219016659894352, | |
| "grad_norm": 1.3713555335998535, | |
| "learning_rate": 4.2991391568918825e-05, | |
| "loss": 1.6675, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.23527021535961, | |
| "grad_norm": 1.406441330909729, | |
| "learning_rate": 4.281282191348289e-05, | |
| "loss": 1.6328, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2515237708248679, | |
| "grad_norm": 1.3209528923034668, | |
| "learning_rate": 4.2632386944129707e-05, | |
| "loss": 1.7064, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.267777326290126, | |
| "grad_norm": 1.4870327711105347, | |
| "learning_rate": 4.245010555555554e-05, | |
| "loss": 1.7161, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.284030881755384, | |
| "grad_norm": 1.4530566930770874, | |
| "learning_rate": 4.2265996835809016e-05, | |
| "loss": 1.6267, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.300284437220642, | |
| "grad_norm": 1.4400850534439087, | |
| "learning_rate": 4.2080080064292304e-05, | |
| "loss": 1.6029, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1535, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.063727570938757e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |