| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9458749343142406, | |
| "eval_steps": 500, | |
| "global_step": 4500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.4763257575757574e-05, | |
| "loss": 4.136, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 2.450021043771044e-05, | |
| "loss": 3.8037, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 2.42371632996633e-05, | |
| "loss": 3.7217, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.3974116161616165e-05, | |
| "loss": 3.7068, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.3711069023569023e-05, | |
| "loss": 3.6665, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.3448021885521888e-05, | |
| "loss": 3.6629, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 2.318497474747475e-05, | |
| "loss": 3.6539, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 2.292192760942761e-05, | |
| "loss": 3.6554, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 2.2658880471380472e-05, | |
| "loss": 3.6323, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 2.2395833333333337e-05, | |
| "loss": 3.6276, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.2132786195286195e-05, | |
| "loss": 3.5915, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2.186973905723906e-05, | |
| "loss": 3.6035, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 2.160669191919192e-05, | |
| "loss": 3.5998, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2.1343644781144782e-05, | |
| "loss": 3.5809, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 2.1080597643097644e-05, | |
| "loss": 3.5888, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 2.0817550505050505e-05, | |
| "loss": 3.6076, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 2.055450336700337e-05, | |
| "loss": 3.5806, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 2.0291456228956228e-05, | |
| "loss": 3.5683, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 2.0028409090909093e-05, | |
| "loss": 3.5725, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 1.9765361952861954e-05, | |
| "loss": 3.57, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 1.9502314814814815e-05, | |
| "loss": 3.5478, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 1.9239267676767677e-05, | |
| "loss": 3.5495, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 1.897622053872054e-05, | |
| "loss": 3.5462, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 1.87131734006734e-05, | |
| "loss": 3.5513, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 1.8450126262626264e-05, | |
| "loss": 3.5378, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 1.8187079124579126e-05, | |
| "loss": 3.5639, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 1.7924031986531987e-05, | |
| "loss": 3.5629, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 1.766098484848485e-05, | |
| "loss": 3.5553, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 1.7397937710437713e-05, | |
| "loss": 3.504, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 1.7134890572390575e-05, | |
| "loss": 3.5303, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.6871843434343436e-05, | |
| "loss": 3.5099, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 1.6608796296296297e-05, | |
| "loss": 3.5221, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 1.634574915824916e-05, | |
| "loss": 3.5422, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 1.608270202020202e-05, | |
| "loss": 3.5353, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 1.581965488215488e-05, | |
| "loss": 3.5026, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 1.5556607744107746e-05, | |
| "loss": 3.5182, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 1.5293560606060604e-05, | |
| "loss": 3.4875, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 1.503051346801347e-05, | |
| "loss": 3.4945, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 1.476746632996633e-05, | |
| "loss": 3.4714, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 1.4504419191919192e-05, | |
| "loss": 3.5158, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 1.4241372053872053e-05, | |
| "loss": 3.5165, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 1.3978324915824916e-05, | |
| "loss": 3.5131, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 1.371527777777778e-05, | |
| "loss": 3.5227, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 1.345223063973064e-05, | |
| "loss": 3.5369, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 1.3189183501683502e-05, | |
| "loss": 3.5149, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 1.2926136363636365e-05, | |
| "loss": 3.4936, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 1.2663089225589225e-05, | |
| "loss": 3.4927, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 1.2400042087542088e-05, | |
| "loss": 3.5001, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 1.213699494949495e-05, | |
| "loss": 3.5218, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 1.1873947811447813e-05, | |
| "loss": 3.4894, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 1.1610900673400674e-05, | |
| "loss": 3.4919, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 1.1347853535353537e-05, | |
| "loss": 3.4681, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 1.1084806397306398e-05, | |
| "loss": 3.4746, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 1.082175925925926e-05, | |
| "loss": 3.4908, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 1.0558712121212121e-05, | |
| "loss": 3.5006, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 1.0295664983164983e-05, | |
| "loss": 3.4835, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 1.0032617845117846e-05, | |
| "loss": 3.4842, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 9.769570707070707e-06, | |
| "loss": 3.4779, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 9.506523569023568e-06, | |
| "loss": 3.4768, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 9.243476430976432e-06, | |
| "loss": 3.4755, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 8.980429292929293e-06, | |
| "loss": 3.4849, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 8.717382154882154e-06, | |
| "loss": 3.481, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 8.454335016835017e-06, | |
| "loss": 3.4918, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 8.191287878787879e-06, | |
| "loss": 3.4624, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 7.928240740740742e-06, | |
| "loss": 3.4826, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 7.665193602693603e-06, | |
| "loss": 3.4719, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 7.402146464646465e-06, | |
| "loss": 3.4858, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 7.139099326599327e-06, | |
| "loss": 3.4616, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 6.876052188552188e-06, | |
| "loss": 3.4697, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 6.613005050505051e-06, | |
| "loss": 3.4486, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 6.349957912457913e-06, | |
| "loss": 3.4718, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 6.086910774410775e-06, | |
| "loss": 3.4443, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 5.823863636363636e-06, | |
| "loss": 3.4849, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5.560816498316499e-06, | |
| "loss": 3.4729, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 5.297769360269361e-06, | |
| "loss": 3.4664, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.034722222222222e-06, | |
| "loss": 3.4643, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 4.7716750841750845e-06, | |
| "loss": 3.4514, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.508627946127946e-06, | |
| "loss": 3.4661, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.245580808080808e-06, | |
| "loss": 3.4637, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.98253367003367e-06, | |
| "loss": 3.453, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.7194865319865326e-06, | |
| "loss": 3.4671, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.456439393939394e-06, | |
| "loss": 3.4439, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 3.1933922558922558e-06, | |
| "loss": 3.4504, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.930345117845118e-06, | |
| "loss": 3.4759, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 2.66729797979798e-06, | |
| "loss": 3.4356, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 2.4042508417508416e-06, | |
| "loss": 3.4618, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 2.141203703703704e-06, | |
| "loss": 3.463, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 1.8781565656565657e-06, | |
| "loss": 3.4524, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 1.6151094276094277e-06, | |
| "loss": 3.4321, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 1.3520622895622895e-06, | |
| "loss": 3.4518, | |
| "step": 4500 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 4757, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 1.104838456586335e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |