| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.6386263390044107, | |
| "global_step": 335000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.921235034656585e-05, | |
| "loss": 0.3334, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.8424700693131696e-05, | |
| "loss": 0.3387, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 1.7637051039697544e-05, | |
| "loss": 0.3327, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.684940138626339e-05, | |
| "loss": 0.3492, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 1.606175173282924e-05, | |
| "loss": 0.3349, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_accuracy": 0.9155995845794678, | |
| "eval_loss": 0.389266699552536, | |
| "eval_runtime": 551.1932, | |
| "eval_samples_per_second": 51.289, | |
| "eval_steps_per_second": 12.823, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 1.5274102079395087e-05, | |
| "loss": 0.3279, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 1.4486452425960932e-05, | |
| "loss": 0.3301, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 1.369880277252678e-05, | |
| "loss": 0.3243, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 1.2911153119092628e-05, | |
| "loss": 0.293, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 1.2123503465658477e-05, | |
| "loss": 0.3053, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "eval_accuracy": 0.9235231876373291, | |
| "eval_loss": 0.3810465931892395, | |
| "eval_runtime": 542.4272, | |
| "eval_samples_per_second": 52.118, | |
| "eval_steps_per_second": 13.03, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 1.1335853812224324e-05, | |
| "loss": 0.3126, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 1.0548204158790173e-05, | |
| "loss": 0.3072, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 9.760554505356018e-06, | |
| "loss": 0.2957, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 8.972904851921865e-06, | |
| "loss": 0.2968, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 8.185255198487714e-06, | |
| "loss": 0.2882, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "eval_accuracy": 0.9224973320960999, | |
| "eval_loss": 0.37537074089050293, | |
| "eval_runtime": 521.9317, | |
| "eval_samples_per_second": 54.164, | |
| "eval_steps_per_second": 13.542, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 7.3976055450535615e-06, | |
| "loss": 0.2754, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 6.6099558916194085e-06, | |
| "loss": 0.2607, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 5.8223062381852555e-06, | |
| "loss": 0.2818, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 5.034656584751103e-06, | |
| "loss": 0.2736, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 4.24700693131695e-06, | |
| "loss": 0.2644, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "eval_accuracy": 0.9297842383384705, | |
| "eval_loss": 0.3645715117454529, | |
| "eval_runtime": 521.9055, | |
| "eval_samples_per_second": 54.167, | |
| "eval_steps_per_second": 13.543, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 3.459357277882798e-06, | |
| "loss": 0.2552, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.6717076244486457e-06, | |
| "loss": 0.266, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.884057971014493e-06, | |
| "loss": 0.2684, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 1.0964083175803404e-06, | |
| "loss": 0.2501, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.087586641461878e-07, | |
| "loss": 0.273, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "eval_accuracy": 0.9299964904785156, | |
| "eval_loss": 0.3369257152080536, | |
| "eval_runtime": 522.7551, | |
| "eval_samples_per_second": 54.079, | |
| "eval_steps_per_second": 13.521, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 1.7952110901071204e-05, | |
| "loss": 0.2834, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 1.787334593572779e-05, | |
| "loss": 0.3047, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 1.7794580970384373e-05, | |
| "loss": 0.2963, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 1.771581600504096e-05, | |
| "loss": 0.3031, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 1.7637051039697544e-05, | |
| "loss": 0.3033, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "eval_accuracy": 0.9257162809371948, | |
| "eval_loss": 0.4006378650665283, | |
| "eval_runtime": 519.4649, | |
| "eval_samples_per_second": 54.421, | |
| "eval_steps_per_second": 13.606, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.755828607435413e-05, | |
| "loss": 0.3024, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 1.7479521109010713e-05, | |
| "loss": 0.3135, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 1.74007561436673e-05, | |
| "loss": 0.3137, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 1.732199117832388e-05, | |
| "loss": 0.3227, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 1.7243226212980467e-05, | |
| "loss": 0.3246, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "eval_accuracy": 0.924018383026123, | |
| "eval_loss": 0.3924681842327118, | |
| "eval_runtime": 518.8244, | |
| "eval_samples_per_second": 54.489, | |
| "eval_steps_per_second": 13.623, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 1.7164461247637053e-05, | |
| "loss": 0.3281, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 1.708569628229364e-05, | |
| "loss": 0.3256, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 1.700693131695022e-05, | |
| "loss": 0.313, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 1.6928166351606807e-05, | |
| "loss": 0.3313, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 1.684940138626339e-05, | |
| "loss": 0.2953, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "eval_accuracy": 0.9212592840194702, | |
| "eval_loss": 0.3895967900753021, | |
| "eval_runtime": 526.2623, | |
| "eval_samples_per_second": 53.718, | |
| "eval_steps_per_second": 13.431, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 1.6770636420919976e-05, | |
| "loss": 0.3103, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 1.669187145557656e-05, | |
| "loss": 0.3089, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 1.6613106490233147e-05, | |
| "loss": 0.3095, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 1.653434152488973e-05, | |
| "loss": 0.3288, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 1.6455576559546316e-05, | |
| "loss": 0.3199, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "eval_accuracy": 0.9203749299049377, | |
| "eval_loss": 0.3942428529262543, | |
| "eval_runtime": 520.6801, | |
| "eval_samples_per_second": 54.294, | |
| "eval_steps_per_second": 13.575, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 1.6376811594202898e-05, | |
| "loss": 0.306, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 1.6298046628859484e-05, | |
| "loss": 0.3104, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 1.621928166351607e-05, | |
| "loss": 0.3139, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 1.6140516698172656e-05, | |
| "loss": 0.3179, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 1.606175173282924e-05, | |
| "loss": 0.3226, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "eval_accuracy": 0.9243367314338684, | |
| "eval_loss": 0.4058537185192108, | |
| "eval_runtime": 552.8946, | |
| "eval_samples_per_second": 51.131, | |
| "eval_steps_per_second": 12.784, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 1.5982986767485824e-05, | |
| "loss": 0.3167, | |
| "step": 255000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 1.5904221802142407e-05, | |
| "loss": 0.3034, | |
| "step": 260000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 1.5825456836798993e-05, | |
| "loss": 0.2976, | |
| "step": 265000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 1.574669187145558e-05, | |
| "loss": 0.3039, | |
| "step": 270000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 1.5667926906112164e-05, | |
| "loss": 0.2889, | |
| "step": 275000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_accuracy": 0.9221436381340027, | |
| "eval_loss": 0.39818692207336426, | |
| "eval_runtime": 508.9512, | |
| "eval_samples_per_second": 55.546, | |
| "eval_steps_per_second": 13.887, | |
| "step": 275000 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 1.5589161940768747e-05, | |
| "loss": 0.3079, | |
| "step": 280000 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 1.5510396975425333e-05, | |
| "loss": 0.3148, | |
| "step": 285000 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 1.5431632010081915e-05, | |
| "loss": 0.2829, | |
| "step": 290000 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 1.53528670447385e-05, | |
| "loss": 0.2978, | |
| "step": 295000 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 1.5274102079395087e-05, | |
| "loss": 0.2963, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "eval_accuracy": 0.922957181930542, | |
| "eval_loss": 0.43214836716651917, | |
| "eval_runtime": 528.9348, | |
| "eval_samples_per_second": 53.447, | |
| "eval_steps_per_second": 13.363, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.5195337114051671e-05, | |
| "loss": 0.3007, | |
| "step": 305000 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.5116572148708256e-05, | |
| "loss": 0.2901, | |
| "step": 310000 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 1.5037807183364841e-05, | |
| "loss": 0.2905, | |
| "step": 315000 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 1.4959042218021424e-05, | |
| "loss": 0.3082, | |
| "step": 320000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 1.488027725267801e-05, | |
| "loss": 0.2899, | |
| "step": 325000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_accuracy": 0.924124538898468, | |
| "eval_loss": 0.40452826023101807, | |
| "eval_runtime": 520.2943, | |
| "eval_samples_per_second": 54.335, | |
| "eval_steps_per_second": 13.585, | |
| "step": 325000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 1.4801512287334594e-05, | |
| "loss": 0.301, | |
| "step": 330000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 1.472274732199118e-05, | |
| "loss": 0.3014, | |
| "step": 335000 | |
| } | |
| ], | |
| "max_steps": 1269600, | |
| "num_train_epochs": 10, | |
| "total_flos": 3.525740928655196e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |