|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 4.995004995004995, |
|
"eval_steps": 500, |
|
"global_step": 5000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 0.000588, |
|
"loss": 0.1326, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"learning_rate": 0.0005759999999999999, |
|
"loss": 0.1179, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"learning_rate": 0.0005639999999999999, |
|
"loss": 0.1292, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 0.000552, |
|
"loss": 0.1292, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"learning_rate": 0.00054, |
|
"loss": 0.1189, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"learning_rate": 0.0005279999999999999, |
|
"loss": 0.0963, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"learning_rate": 0.000516, |
|
"loss": 0.0949, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"learning_rate": 0.0005039999999999999, |
|
"loss": 0.1309, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"learning_rate": 0.0004919999999999999, |
|
"loss": 0.1353, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"learning_rate": 0.00047999999999999996, |
|
"loss": 0.1306, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy_background": NaN, |
|
"eval_accuracy_crack": 0.0, |
|
"eval_iou_background": 0.0, |
|
"eval_iou_crack": 0.0, |
|
"eval_loss": 0.11144199222326279, |
|
"eval_mean_accuracy": 0.0, |
|
"eval_mean_iou": 0.0, |
|
"eval_overall_accuracy": 0.0, |
|
"eval_runtime": 523.6071, |
|
"eval_samples_per_second": 0.338, |
|
"eval_steps_per_second": 0.338, |
|
"step": 1001 |
|
}, |
|
{ |
|
"epoch": 1.1, |
|
"learning_rate": 0.000468, |
|
"loss": 0.113, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"learning_rate": 0.00045599999999999997, |
|
"loss": 0.1031, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 1.3, |
|
"learning_rate": 0.00044399999999999995, |
|
"loss": 0.1021, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 1.4, |
|
"learning_rate": 0.00043199999999999993, |
|
"loss": 0.1107, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 1.5, |
|
"learning_rate": 0.00041999999999999996, |
|
"loss": 0.0944, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"learning_rate": 0.000408, |
|
"loss": 0.1082, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 1.7, |
|
"learning_rate": 0.000396, |
|
"loss": 0.0868, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 1.8, |
|
"learning_rate": 0.00038399999999999996, |
|
"loss": 0.1218, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 1.9, |
|
"learning_rate": 0.000372, |
|
"loss": 0.1062, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"learning_rate": 0.00035999999999999997, |
|
"loss": 0.107, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy_background": NaN, |
|
"eval_accuracy_crack": 5.917486567305492e-06, |
|
"eval_iou_background": 0.0, |
|
"eval_iou_crack": 5.917486567305492e-06, |
|
"eval_loss": 0.12381087243556976, |
|
"eval_mean_accuracy": 5.917486567305492e-06, |
|
"eval_mean_iou": 2.958743283652746e-06, |
|
"eval_overall_accuracy": 5.917486567305492e-06, |
|
"eval_runtime": 527.1442, |
|
"eval_samples_per_second": 0.336, |
|
"eval_steps_per_second": 0.336, |
|
"step": 2002 |
|
}, |
|
{ |
|
"epoch": 2.1, |
|
"learning_rate": 0.00034799999999999995, |
|
"loss": 0.0988, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 2.2, |
|
"learning_rate": 0.000336, |
|
"loss": 0.0966, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 2.3, |
|
"learning_rate": 0.000324, |
|
"loss": 0.1167, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"learning_rate": 0.000312, |
|
"loss": 0.088, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 2.5, |
|
"learning_rate": 0.0003, |
|
"loss": 0.0806, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 2.6, |
|
"learning_rate": 0.00028799999999999995, |
|
"loss": 0.0864, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 2.7, |
|
"learning_rate": 0.000276, |
|
"loss": 0.102, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"learning_rate": 0.00026399999999999997, |
|
"loss": 0.0957, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 2.9, |
|
"learning_rate": 0.00025199999999999995, |
|
"loss": 0.111, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"learning_rate": 0.00023999999999999998, |
|
"loss": 0.1285, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy_background": NaN, |
|
"eval_accuracy_crack": 0.004886857656833119, |
|
"eval_iou_background": 0.0, |
|
"eval_iou_crack": 0.004827905390592649, |
|
"eval_loss": 0.1630975902080536, |
|
"eval_mean_accuracy": 0.004886857656833119, |
|
"eval_mean_iou": 0.0024139526952963244, |
|
"eval_overall_accuracy": 0.004886857656833119, |
|
"eval_runtime": 529.3979, |
|
"eval_samples_per_second": 0.334, |
|
"eval_steps_per_second": 0.334, |
|
"step": 3003 |
|
}, |
|
{ |
|
"epoch": 3.1, |
|
"learning_rate": 0.00022799999999999999, |
|
"loss": 0.0961, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 3.2, |
|
"learning_rate": 0.00021599999999999996, |
|
"loss": 0.0937, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 3.3, |
|
"learning_rate": 0.000204, |
|
"loss": 0.1059, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 3.4, |
|
"learning_rate": 0.00019199999999999998, |
|
"loss": 0.1069, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 3.5, |
|
"learning_rate": 0.00017999999999999998, |
|
"loss": 0.0571, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 3.6, |
|
"learning_rate": 0.000168, |
|
"loss": 0.0958, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 3.7, |
|
"learning_rate": 0.000156, |
|
"loss": 0.0836, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 3.8, |
|
"learning_rate": 0.00014399999999999998, |
|
"loss": 0.0988, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 3.9, |
|
"learning_rate": 0.00013199999999999998, |
|
"loss": 0.0675, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"learning_rate": 0.00011999999999999999, |
|
"loss": 0.0887, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy_background": NaN, |
|
"eval_accuracy_crack": 0.0003008055671713625, |
|
"eval_iou_background": 0.0, |
|
"eval_iou_crack": 0.0003008055671713625, |
|
"eval_loss": 0.10831139981746674, |
|
"eval_mean_accuracy": 0.0003008055671713625, |
|
"eval_mean_iou": 0.00015040278358568126, |
|
"eval_overall_accuracy": 0.0003008055671713625, |
|
"eval_runtime": 530.7687, |
|
"eval_samples_per_second": 0.333, |
|
"eval_steps_per_second": 0.333, |
|
"step": 4004 |
|
}, |
|
{ |
|
"epoch": 4.1, |
|
"learning_rate": 0.00010799999999999998, |
|
"loss": 0.075, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 4.2, |
|
"learning_rate": 9.599999999999999e-05, |
|
"loss": 0.0891, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 4.3, |
|
"learning_rate": 8.4e-05, |
|
"loss": 0.0745, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 4.4, |
|
"learning_rate": 7.199999999999999e-05, |
|
"loss": 0.092, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 4.5, |
|
"learning_rate": 5.9999999999999995e-05, |
|
"loss": 0.0706, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 4.6, |
|
"learning_rate": 4.7999999999999994e-05, |
|
"loss": 0.0957, |
|
"step": 4600 |
|
}, |
|
{ |
|
"epoch": 4.7, |
|
"learning_rate": 3.5999999999999994e-05, |
|
"loss": 0.0776, |
|
"step": 4700 |
|
}, |
|
{ |
|
"epoch": 4.8, |
|
"learning_rate": 2.3999999999999997e-05, |
|
"loss": 0.0895, |
|
"step": 4800 |
|
}, |
|
{ |
|
"epoch": 4.9, |
|
"learning_rate": 1.1999999999999999e-05, |
|
"loss": 0.0795, |
|
"step": 4900 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"learning_rate": 0.0, |
|
"loss": 0.0828, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy_background": NaN, |
|
"eval_accuracy_crack": 0.0009813165224114941, |
|
"eval_iou_background": 0.0, |
|
"eval_iou_crack": 0.0009813165224114941, |
|
"eval_loss": 0.1343701034784317, |
|
"eval_mean_accuracy": 0.0009813165224114941, |
|
"eval_mean_iou": 0.0004906582612057471, |
|
"eval_overall_accuracy": 0.0009813165224114941, |
|
"eval_runtime": 531.91, |
|
"eval_samples_per_second": 0.333, |
|
"eval_steps_per_second": 0.333, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"step": 5000, |
|
"total_flos": 1.99584366723072e+18, |
|
"train_loss": 0.09988212461471557, |
|
"train_runtime": 34878.9562, |
|
"train_samples_per_second": 0.143, |
|
"train_steps_per_second": 0.143 |
|
} |
|
], |
|
"logging_steps": 100, |
|
"max_steps": 5000, |
|
"num_train_epochs": 5, |
|
"save_steps": 500, |
|
"total_flos": 1.99584366723072e+18, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|