|
{ |
|
"best_metric": 0.06634166091680527, |
|
"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": 1.5509302616119385, |
|
"learning_rate": 0.00019230769230769233, |
|
"loss": 0.7749, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.3076923076923077, |
|
"grad_norm": 3.293163299560547, |
|
"learning_rate": 0.00018461538461538463, |
|
"loss": 0.2869, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.46153846153846156, |
|
"grad_norm": 0.553378701210022, |
|
"learning_rate": 0.00017692307692307693, |
|
"loss": 0.2031, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.6153846153846154, |
|
"grad_norm": 1.476561427116394, |
|
"learning_rate": 0.00016923076923076923, |
|
"loss": 0.2127, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.7692307692307693, |
|
"grad_norm": 1.6264675855636597, |
|
"learning_rate": 0.00016153846153846155, |
|
"loss": 0.1767, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.9230769230769231, |
|
"grad_norm": 3.1017961502075195, |
|
"learning_rate": 0.00015384615384615385, |
|
"loss": 0.2664, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 1.0769230769230769, |
|
"grad_norm": 0.2297724187374115, |
|
"learning_rate": 0.00014615384615384615, |
|
"loss": 0.095, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 1.2307692307692308, |
|
"grad_norm": 0.13144992291927338, |
|
"learning_rate": 0.00013846153846153847, |
|
"loss": 0.1187, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 1.3846153846153846, |
|
"grad_norm": 0.14720702171325684, |
|
"learning_rate": 0.00013076923076923077, |
|
"loss": 0.0459, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 1.5384615384615383, |
|
"grad_norm": 2.3203179836273193, |
|
"learning_rate": 0.0001230769230769231, |
|
"loss": 0.0997, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 1.5384615384615383, |
|
"eval_accuracy": 0.9774436090225563, |
|
"eval_loss": 0.08642897754907608, |
|
"eval_runtime": 1.3472, |
|
"eval_samples_per_second": 98.724, |
|
"eval_steps_per_second": 12.619, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 1.6923076923076923, |
|
"grad_norm": 1.8982298374176025, |
|
"learning_rate": 0.00011538461538461538, |
|
"loss": 0.1221, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 1.8461538461538463, |
|
"grad_norm": 1.9391183853149414, |
|
"learning_rate": 0.0001076923076923077, |
|
"loss": 0.1486, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 0.1778109073638916, |
|
"learning_rate": 0.0001, |
|
"loss": 0.0905, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 2.1538461538461537, |
|
"grad_norm": 0.08789520710706711, |
|
"learning_rate": 9.230769230769232e-05, |
|
"loss": 0.0415, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 2.3076923076923075, |
|
"grad_norm": 0.12370666861534119, |
|
"learning_rate": 8.461538461538461e-05, |
|
"loss": 0.054, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 2.4615384615384617, |
|
"grad_norm": 0.4401356875896454, |
|
"learning_rate": 7.692307692307693e-05, |
|
"loss": 0.033, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 2.6153846153846154, |
|
"grad_norm": 0.0808761939406395, |
|
"learning_rate": 6.923076923076924e-05, |
|
"loss": 0.0299, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 2.769230769230769, |
|
"grad_norm": 10.307175636291504, |
|
"learning_rate": 6.153846153846155e-05, |
|
"loss": 0.0259, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 2.9230769230769234, |
|
"grad_norm": 1.5089224576950073, |
|
"learning_rate": 5.384615384615385e-05, |
|
"loss": 0.069, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 3.076923076923077, |
|
"grad_norm": 0.06397894024848938, |
|
"learning_rate": 4.615384615384616e-05, |
|
"loss": 0.0125, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 3.076923076923077, |
|
"eval_accuracy": 0.9774436090225563, |
|
"eval_loss": 0.06634166091680527, |
|
"eval_runtime": 1.2786, |
|
"eval_samples_per_second": 104.016, |
|
"eval_steps_per_second": 13.295, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 3.230769230769231, |
|
"grad_norm": 0.06174139305949211, |
|
"learning_rate": 3.846153846153846e-05, |
|
"loss": 0.0122, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 3.3846153846153846, |
|
"grad_norm": 0.0587957464158535, |
|
"learning_rate": 3.0769230769230774e-05, |
|
"loss": 0.0123, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 3.5384615384615383, |
|
"grad_norm": 0.06546800583600998, |
|
"learning_rate": 2.307692307692308e-05, |
|
"loss": 0.0111, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 3.6923076923076925, |
|
"grad_norm": 0.0588623583316803, |
|
"learning_rate": 1.5384615384615387e-05, |
|
"loss": 0.0112, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 3.8461538461538463, |
|
"grad_norm": 0.05679621547460556, |
|
"learning_rate": 7.692307692307694e-06, |
|
"loss": 0.0109, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"grad_norm": 0.05850306525826454, |
|
"learning_rate": 0.0, |
|
"loss": 0.0358, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"step": 260, |
|
"total_flos": 3.205097416476426e+17, |
|
"train_loss": 0.11540300791653303, |
|
"train_runtime": 92.7413, |
|
"train_samples_per_second": 44.597, |
|
"train_steps_per_second": 2.803 |
|
} |
|
], |
|
"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": false |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 3.205097416476426e+17, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|