{ "best_metric": 0.0006729721208103001, "best_model_checkpoint": "./vit-base-blur/checkpoint-400", "epoch": 10.0, "global_step": 440, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.23, "learning_rate": 0.00019545454545454548, "loss": 0.0392, "step": 10 }, { "epoch": 0.45, "learning_rate": 0.00019090909090909092, "loss": 0.1092, "step": 20 }, { "epoch": 0.68, "learning_rate": 0.00018636363636363636, "loss": 0.064, "step": 30 }, { "epoch": 0.91, "learning_rate": 0.00018181818181818183, "loss": 0.0254, "step": 40 }, { "epoch": 1.14, "learning_rate": 0.00017727272727272728, "loss": 0.0662, "step": 50 }, { "epoch": 1.36, "learning_rate": 0.00017272727272727275, "loss": 0.0306, "step": 60 }, { "epoch": 1.59, "learning_rate": 0.0001681818181818182, "loss": 0.0073, "step": 70 }, { "epoch": 1.82, "learning_rate": 0.00016363636363636366, "loss": 0.0231, "step": 80 }, { "epoch": 2.05, "learning_rate": 0.0001590909090909091, "loss": 0.0029, "step": 90 }, { "epoch": 2.27, "learning_rate": 0.00015454545454545454, "loss": 0.0027, "step": 100 }, { "epoch": 2.27, "eval_accuracy": 0.987012987012987, "eval_loss": 0.05653372034430504, "eval_runtime": 2.757, "eval_samples_per_second": 27.928, "eval_steps_per_second": 3.627, "step": 100 }, { "epoch": 2.5, "learning_rate": 0.00015000000000000001, "loss": 0.0022, "step": 110 }, { "epoch": 2.73, "learning_rate": 0.00014545454545454546, "loss": 0.0226, "step": 120 }, { "epoch": 2.95, "learning_rate": 0.00014090909090909093, "loss": 0.02, "step": 130 }, { "epoch": 3.18, "learning_rate": 0.00013636363636363637, "loss": 0.0137, "step": 140 }, { "epoch": 3.41, "learning_rate": 0.0001318181818181818, "loss": 0.0016, "step": 150 }, { "epoch": 3.64, "learning_rate": 0.00012727272727272728, "loss": 0.0016, "step": 160 }, { "epoch": 3.86, "learning_rate": 0.00012272727272727272, "loss": 0.037, "step": 170 }, { "epoch": 4.09, "learning_rate": 0.0001181818181818182, "loss": 0.0016, "step": 180 }, { "epoch": 4.32, "learning_rate": 0.00011363636363636365, "loss": 0.0041, "step": 190 }, { "epoch": 4.55, "learning_rate": 0.00010909090909090909, "loss": 0.0013, "step": 200 }, { "epoch": 4.55, "eval_accuracy": 1.0, "eval_loss": 0.001206980086863041, "eval_runtime": 2.4286, "eval_samples_per_second": 31.706, "eval_steps_per_second": 4.118, "step": 200 }, { "epoch": 4.77, "learning_rate": 0.00010454545454545455, "loss": 0.0012, "step": 210 }, { "epoch": 5.0, "learning_rate": 0.0001, "loss": 0.0011, "step": 220 }, { "epoch": 5.23, "learning_rate": 9.545454545454546e-05, "loss": 0.0011, "step": 230 }, { "epoch": 5.45, "learning_rate": 9.090909090909092e-05, "loss": 0.001, "step": 240 }, { "epoch": 5.68, "learning_rate": 8.636363636363637e-05, "loss": 0.001, "step": 250 }, { "epoch": 5.91, "learning_rate": 8.181818181818183e-05, "loss": 0.0009, "step": 260 }, { "epoch": 6.14, "learning_rate": 7.727272727272727e-05, "loss": 0.0009, "step": 270 }, { "epoch": 6.36, "learning_rate": 7.272727272727273e-05, "loss": 0.0009, "step": 280 }, { "epoch": 6.59, "learning_rate": 6.818181818181818e-05, "loss": 0.0008, "step": 290 }, { "epoch": 6.82, "learning_rate": 6.363636363636364e-05, "loss": 0.0008, "step": 300 }, { "epoch": 6.82, "eval_accuracy": 1.0, "eval_loss": 0.0008008014992810786, "eval_runtime": 1.9546, "eval_samples_per_second": 39.395, "eval_steps_per_second": 5.116, "step": 300 }, { "epoch": 7.05, "learning_rate": 5.90909090909091e-05, "loss": 0.0008, "step": 310 }, { "epoch": 7.27, "learning_rate": 5.4545454545454546e-05, "loss": 0.0008, "step": 320 }, { "epoch": 7.5, "learning_rate": 5e-05, "loss": 0.0007, "step": 330 }, { "epoch": 7.73, "learning_rate": 4.545454545454546e-05, "loss": 0.0007, "step": 340 }, { "epoch": 7.95, "learning_rate": 4.0909090909090915e-05, "loss": 0.0007, "step": 350 }, { "epoch": 8.18, "learning_rate": 3.6363636363636364e-05, "loss": 0.0007, "step": 360 }, { "epoch": 8.41, "learning_rate": 3.181818181818182e-05, "loss": 0.0007, "step": 370 }, { "epoch": 8.64, "learning_rate": 2.7272727272727273e-05, "loss": 0.0007, "step": 380 }, { "epoch": 8.86, "learning_rate": 2.272727272727273e-05, "loss": 0.0007, "step": 390 }, { "epoch": 9.09, "learning_rate": 1.8181818181818182e-05, "loss": 0.0007, "step": 400 }, { "epoch": 9.09, "eval_accuracy": 1.0, "eval_loss": 0.0006729721208103001, "eval_runtime": 1.9389, "eval_samples_per_second": 39.714, "eval_steps_per_second": 5.158, "step": 400 }, { "epoch": 9.32, "learning_rate": 1.3636363636363637e-05, "loss": 0.0007, "step": 410 }, { "epoch": 9.55, "learning_rate": 9.090909090909091e-06, "loss": 0.0007, "step": 420 }, { "epoch": 9.77, "learning_rate": 4.5454545454545455e-06, "loss": 0.0007, "step": 430 }, { "epoch": 10.0, "learning_rate": 0.0, "loss": 0.0007, "step": 440 }, { "epoch": 10.0, "step": 440, "total_flos": 5.354696482368307e+17, "train_loss": 0.011267724910496988, "train_runtime": 422.1478, "train_samples_per_second": 16.369, "train_steps_per_second": 1.042 } ], "max_steps": 440, "num_train_epochs": 10, "total_flos": 5.354696482368307e+17, "trial_name": null, "trial_params": null }