| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 29370, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17024174327545114, | |
| "grad_norm": 0.6650531888008118, | |
| "learning_rate": 1.96595165134491e-05, | |
| "loss": 0.2305, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.34048348655090227, | |
| "grad_norm": 2.3951523303985596, | |
| "learning_rate": 1.9319033026898198e-05, | |
| "loss": 0.1633, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5107252298263534, | |
| "grad_norm": 0.9463224411010742, | |
| "learning_rate": 1.8978549540347296e-05, | |
| "loss": 0.1548, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6809669731018045, | |
| "grad_norm": 1.2913334369659424, | |
| "learning_rate": 1.8638066053796395e-05, | |
| "loss": 0.1487, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8512087163772557, | |
| "grad_norm": 1.0490585565567017, | |
| "learning_rate": 1.829758256724549e-05, | |
| "loss": 0.1435, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.0214504596527068, | |
| "grad_norm": 1.5652408599853516, | |
| "learning_rate": 1.7957099080694588e-05, | |
| "loss": 0.1397, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.191692202928158, | |
| "grad_norm": 1.1205641031265259, | |
| "learning_rate": 1.7616615594143686e-05, | |
| "loss": 0.1283, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.361933946203609, | |
| "grad_norm": 1.0744216442108154, | |
| "learning_rate": 1.727613210759278e-05, | |
| "loss": 0.1292, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.5321756894790601, | |
| "grad_norm": 1.089113712310791, | |
| "learning_rate": 1.693564862104188e-05, | |
| "loss": 0.1273, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.7024174327545114, | |
| "grad_norm": 2.334705114364624, | |
| "learning_rate": 1.6595165134490977e-05, | |
| "loss": 0.1275, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.8726591760299627, | |
| "grad_norm": 1.1323754787445068, | |
| "learning_rate": 1.6254681647940076e-05, | |
| "loss": 0.1251, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.0429009193054135, | |
| "grad_norm": 0.8757261633872986, | |
| "learning_rate": 1.5914198161389174e-05, | |
| "loss": 0.1213, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.213142662580865, | |
| "grad_norm": 1.1232839822769165, | |
| "learning_rate": 1.5573714674838272e-05, | |
| "loss": 0.1104, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.383384405856316, | |
| "grad_norm": 0.8715490698814392, | |
| "learning_rate": 1.5233231188287369e-05, | |
| "loss": 0.1099, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.553626149131767, | |
| "grad_norm": 1.2656769752502441, | |
| "learning_rate": 1.4892747701736467e-05, | |
| "loss": 0.1102, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.723867892407218, | |
| "grad_norm": 1.1669204235076904, | |
| "learning_rate": 1.4552264215185565e-05, | |
| "loss": 0.1101, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.8941096356826694, | |
| "grad_norm": 1.0073705911636353, | |
| "learning_rate": 1.4211780728634664e-05, | |
| "loss": 0.1085, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.0643513789581207, | |
| "grad_norm": 1.1393821239471436, | |
| "learning_rate": 1.3871297242083762e-05, | |
| "loss": 0.1027, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.2345931222335715, | |
| "grad_norm": 1.4679887294769287, | |
| "learning_rate": 1.3530813755532857e-05, | |
| "loss": 0.0926, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.404834865509023, | |
| "grad_norm": 0.8374710083007812, | |
| "learning_rate": 1.3190330268981955e-05, | |
| "loss": 0.0925, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.575076608784474, | |
| "grad_norm": 1.2514032125473022, | |
| "learning_rate": 1.2849846782431053e-05, | |
| "loss": 0.0927, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.7453183520599254, | |
| "grad_norm": 1.5251351594924927, | |
| "learning_rate": 1.250936329588015e-05, | |
| "loss": 0.0929, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.915560095335376, | |
| "grad_norm": 1.0668872594833374, | |
| "learning_rate": 1.2168879809329248e-05, | |
| "loss": 0.0923, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 4.085801838610827, | |
| "grad_norm": 1.0528796911239624, | |
| "learning_rate": 1.1828396322778346e-05, | |
| "loss": 0.0848, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.256043581886279, | |
| "grad_norm": 1.316041111946106, | |
| "learning_rate": 1.1487912836227445e-05, | |
| "loss": 0.0767, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 4.42628532516173, | |
| "grad_norm": 1.6180927753448486, | |
| "learning_rate": 1.1147429349676541e-05, | |
| "loss": 0.077, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 4.596527068437181, | |
| "grad_norm": 1.2156362533569336, | |
| "learning_rate": 1.080694586312564e-05, | |
| "loss": 0.0773, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 4.766768811712632, | |
| "grad_norm": 1.621887445449829, | |
| "learning_rate": 1.0466462376574738e-05, | |
| "loss": 0.0773, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.937010554988083, | |
| "grad_norm": 1.5306437015533447, | |
| "learning_rate": 1.0125978890023836e-05, | |
| "loss": 0.0774, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 5.107252298263534, | |
| "grad_norm": 22.37914276123047, | |
| "learning_rate": 9.785495403472932e-06, | |
| "loss": 0.0678, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 5.2774940415389855, | |
| "grad_norm": 1.3330860137939453, | |
| "learning_rate": 9.44501191692203e-06, | |
| "loss": 0.0634, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 5.447735784814436, | |
| "grad_norm": 1.9692567586898804, | |
| "learning_rate": 9.104528430371127e-06, | |
| "loss": 0.0634, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 5.617977528089888, | |
| "grad_norm": 1.3089221715927124, | |
| "learning_rate": 8.764044943820226e-06, | |
| "loss": 0.0635, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 5.788219271365339, | |
| "grad_norm": 1.5806821584701538, | |
| "learning_rate": 8.423561457269324e-06, | |
| "loss": 0.0637, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 5.95846101464079, | |
| "grad_norm": 1.579941987991333, | |
| "learning_rate": 8.08307797071842e-06, | |
| "loss": 0.0633, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 6.128702757916241, | |
| "grad_norm": 1.5726784467697144, | |
| "learning_rate": 7.742594484167519e-06, | |
| "loss": 0.054, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 6.298944501191692, | |
| "grad_norm": 1.140791654586792, | |
| "learning_rate": 7.402110997616616e-06, | |
| "loss": 0.052, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 6.469186244467143, | |
| "grad_norm": 1.6548409461975098, | |
| "learning_rate": 7.061627511065714e-06, | |
| "loss": 0.0516, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 6.639427987742595, | |
| "grad_norm": 1.3514069318771362, | |
| "learning_rate": 6.721144024514812e-06, | |
| "loss": 0.0522, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 6.809669731018046, | |
| "grad_norm": 1.5590009689331055, | |
| "learning_rate": 6.38066053796391e-06, | |
| "loss": 0.0518, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 6.9799114742934965, | |
| "grad_norm": 1.2986799478530884, | |
| "learning_rate": 6.0401770514130066e-06, | |
| "loss": 0.0524, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 7.150153217568948, | |
| "grad_norm": 1.5317639112472534, | |
| "learning_rate": 5.699693564862104e-06, | |
| "loss": 0.044, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 7.320394960844399, | |
| "grad_norm": 2.344708204269409, | |
| "learning_rate": 5.359210078311202e-06, | |
| "loss": 0.0415, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 7.49063670411985, | |
| "grad_norm": 3.3057548999786377, | |
| "learning_rate": 5.0187265917603005e-06, | |
| "loss": 0.0418, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 7.6608784473953015, | |
| "grad_norm": 1.3382242918014526, | |
| "learning_rate": 4.678243105209398e-06, | |
| "loss": 0.0419, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 7.831120190670752, | |
| "grad_norm": 1.7018738985061646, | |
| "learning_rate": 4.337759618658495e-06, | |
| "loss": 0.0421, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 8.001361933946203, | |
| "grad_norm": 0.9316732883453369, | |
| "learning_rate": 3.997276132107593e-06, | |
| "loss": 0.0414, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 8.171603677221654, | |
| "grad_norm": 1.4249956607818604, | |
| "learning_rate": 3.656792645556691e-06, | |
| "loss": 0.0346, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 8.341845420497107, | |
| "grad_norm": 1.263279914855957, | |
| "learning_rate": 3.3163091590057884e-06, | |
| "loss": 0.0345, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 8.512087163772557, | |
| "grad_norm": 2.6162939071655273, | |
| "learning_rate": 2.9758256724548862e-06, | |
| "loss": 0.0342, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 8.682328907048008, | |
| "grad_norm": 1.2574002742767334, | |
| "learning_rate": 2.635342185903984e-06, | |
| "loss": 0.0345, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 8.85257065032346, | |
| "grad_norm": 5.4230732917785645, | |
| "learning_rate": 2.2948586993530815e-06, | |
| "loss": 0.0344, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 9.02281239359891, | |
| "grad_norm": 0.885810136795044, | |
| "learning_rate": 1.9543752128021793e-06, | |
| "loss": 0.0333, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 9.19305413687436, | |
| "grad_norm": 1.7516717910766602, | |
| "learning_rate": 1.6138917262512767e-06, | |
| "loss": 0.0291, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 9.363295880149813, | |
| "grad_norm": 1.1372159719467163, | |
| "learning_rate": 1.2734082397003748e-06, | |
| "loss": 0.0293, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 9.533537623425264, | |
| "grad_norm": 0.9269993305206299, | |
| "learning_rate": 9.329247531494723e-07, | |
| "loss": 0.0294, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 9.703779366700715, | |
| "grad_norm": 1.229074239730835, | |
| "learning_rate": 5.9244126659857e-07, | |
| "loss": 0.0291, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 9.874021109976166, | |
| "grad_norm": 2.4099299907684326, | |
| "learning_rate": 2.519577800476677e-07, | |
| "loss": 0.0289, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 29370, | |
| "total_flos": 8.758967154215731e+17, | |
| "train_loss": 0.07915824020562384, | |
| "train_runtime": 29556.1772, | |
| "train_samples_per_second": 31.797, | |
| "train_steps_per_second": 0.994 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 29370, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.758967154215731e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |