| { | |
| "best_metric": 0.06726039946079254, | |
| "best_model_checkpoint": "./vit-base-beans-demo-v18Set/checkpoint-200", | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 260, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 1.554421305656433, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.7831, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.9934561252593994, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.2972, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.38546833395957947, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.1788, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 2.607243537902832, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 0.2693, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 5.0497307777404785, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.2568, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 1.8226447105407715, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.3404, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.3479740023612976, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 0.213, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 1.2378307580947876, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.1413, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.21435503661632538, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.0498, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 2.7997779846191406, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0625, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.9774436090225563, | |
| "eval_loss": 0.10245038568973541, | |
| "eval_runtime": 1.2575, | |
| "eval_samples_per_second": 105.767, | |
| "eval_steps_per_second": 13.519, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 0.26306748390197754, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.1177, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.10240321606397629, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 0.1089, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.7692477703094482, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0774, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 3.356201171875, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.0397, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.0990709736943245, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.0232, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 2.2368693351745605, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0292, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.06961995363235474, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.0241, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.06873491406440735, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.016, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 4.5112223625183105, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.0263, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.060032740235328674, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0119, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 0.9849624060150376, | |
| "eval_loss": 0.06726039946079254, | |
| "eval_runtime": 1.2772, | |
| "eval_samples_per_second": 104.135, | |
| "eval_steps_per_second": 13.31, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.20212945342063904, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.0116, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.05760957673192024, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0113, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.0580044649541378, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0107, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.05357787013053894, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0104, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.05624014511704445, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0103, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.05524070933461189, | |
| "learning_rate": 0.0, | |
| "loss": 0.0103, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.12043425440788269, | |
| "train_runtime": 117.3246, | |
| "train_samples_per_second": 35.253, | |
| "train_steps_per_second": 2.216 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 260, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |