| { | |
| "best_global_step": 200, | |
| "best_metric": 0.9924812030075187, | |
| "best_model_checkpoint": "./vit-base-beans-demo-v5/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": 1.8465008735656738, | |
| "learning_rate": 0.0001930769230769231, | |
| "loss": 0.8272, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.3152756690979004, | |
| "learning_rate": 0.0001853846153846154, | |
| "loss": 0.3076, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 1.462149977684021, | |
| "learning_rate": 0.0001776923076923077, | |
| "loss": 0.2484, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 1.5601600408554077, | |
| "learning_rate": 0.00017, | |
| "loss": 0.3078, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 1.7110002040863037, | |
| "learning_rate": 0.0001623076923076923, | |
| "loss": 0.2375, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 1.6433830261230469, | |
| "learning_rate": 0.00015461538461538464, | |
| "loss": 0.2519, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.5108979940414429, | |
| "learning_rate": 0.00014692307692307693, | |
| "loss": 0.1594, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.17463286221027374, | |
| "learning_rate": 0.00013923076923076923, | |
| "loss": 0.0964, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 6.128848552703857, | |
| "learning_rate": 0.00013153846153846156, | |
| "loss": 0.0696, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.13578061759471893, | |
| "learning_rate": 0.00012384615384615385, | |
| "loss": 0.0814, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.9097744360902256, | |
| "eval_loss": 0.21250881254673004, | |
| "eval_runtime": 1.2493, | |
| "eval_samples_per_second": 106.461, | |
| "eval_steps_per_second": 7.204, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 1.2079147100448608, | |
| "learning_rate": 0.00011615384615384617, | |
| "loss": 0.1222, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 2.105736255645752, | |
| "learning_rate": 0.00010846153846153846, | |
| "loss": 0.0689, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0425434112548828, | |
| "learning_rate": 0.00010076923076923077, | |
| "loss": 0.0689, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.09185847640037537, | |
| "learning_rate": 9.307692307692309e-05, | |
| "loss": 0.0353, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 9.617168426513672, | |
| "learning_rate": 8.538461538461538e-05, | |
| "loss": 0.0381, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.08445341140031815, | |
| "learning_rate": 7.76923076923077e-05, | |
| "loss": 0.017, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.07447629421949387, | |
| "learning_rate": 7e-05, | |
| "loss": 0.0163, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 2.2068941593170166, | |
| "learning_rate": 6.23076923076923e-05, | |
| "loss": 0.0237, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 1.7057480812072754, | |
| "learning_rate": 5.461538461538461e-05, | |
| "loss": 0.0683, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.06541252136230469, | |
| "learning_rate": 4.692307692307693e-05, | |
| "loss": 0.0127, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 0.9924812030075187, | |
| "eval_loss": 0.027433352544903755, | |
| "eval_runtime": 1.2357, | |
| "eval_samples_per_second": 107.631, | |
| "eval_steps_per_second": 7.283, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.06294204294681549, | |
| "learning_rate": 3.923076923076923e-05, | |
| "loss": 0.0124, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.06276626139879227, | |
| "learning_rate": 3.153846153846154e-05, | |
| "loss": 0.0125, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 7.931839466094971, | |
| "learning_rate": 2.384615384615385e-05, | |
| "loss": 0.0306, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.061203066259622574, | |
| "learning_rate": 1.6153846153846154e-05, | |
| "loss": 0.0115, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.07784971594810486, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.0117, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.05536546930670738, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 0.0118, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.12112718929465001, | |
| "train_runtime": 104.2153, | |
| "train_samples_per_second": 39.687, | |
| "train_steps_per_second": 2.495 | |
| } | |
| ], | |
| "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 | |
| } | |