|
{ |
|
"best_metric": 0.08830931037664413, |
|
"best_model_checkpoint": "/home1/datawork/mcontini/models/multilabel/huggingface/dinov2-large-2024_01_14-without_data_aug_batch-size32_epochs20_freeze/checkpoint-5480", |
|
"epoch": 20.0, |
|
"eval_steps": 500, |
|
"global_step": 5480, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.4696441032798325, |
|
"eval_f1_macro": 0.5445473909109761, |
|
"eval_f1_micro": 0.7663056558363417, |
|
"eval_loss": 0.1264539361000061, |
|
"eval_roc_auc": 0.8400878306196953, |
|
"eval_runtime": 1036.6447, |
|
"eval_samples_per_second": 2.765, |
|
"eval_steps_per_second": 0.087, |
|
"learning_rate": 0.001, |
|
"step": 274 |
|
}, |
|
{ |
|
"epoch": 1.82, |
|
"learning_rate": 0.001, |
|
"loss": 0.2337, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.5115143056524774, |
|
"eval_f1_macro": 0.6503987559811281, |
|
"eval_f1_micro": 0.8025544267053701, |
|
"eval_loss": 0.11075052618980408, |
|
"eval_roc_auc": 0.8693871473003357, |
|
"eval_runtime": 1011.3901, |
|
"eval_samples_per_second": 2.834, |
|
"eval_steps_per_second": 0.089, |
|
"learning_rate": 0.001, |
|
"step": 548 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.5177948360083741, |
|
"eval_f1_macro": 0.6963127404974011, |
|
"eval_f1_micro": 0.8184487821730869, |
|
"eval_loss": 0.1064244732260704, |
|
"eval_roc_auc": 0.88043822362814, |
|
"eval_runtime": 1010.6496, |
|
"eval_samples_per_second": 2.836, |
|
"eval_steps_per_second": 0.089, |
|
"learning_rate": 0.001, |
|
"step": 822 |
|
}, |
|
{ |
|
"epoch": 3.65, |
|
"learning_rate": 0.001, |
|
"loss": 0.1259, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.5188415910676901, |
|
"eval_f1_macro": 0.6838420416441195, |
|
"eval_f1_micro": 0.8164364720041263, |
|
"eval_loss": 0.10615076124668121, |
|
"eval_roc_auc": 0.8808295127544562, |
|
"eval_runtime": 1012.7035, |
|
"eval_samples_per_second": 2.83, |
|
"eval_steps_per_second": 0.089, |
|
"learning_rate": 0.001, |
|
"step": 1096 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.4965108164689463, |
|
"eval_f1_macro": 0.6749223162423283, |
|
"eval_f1_micro": 0.8156853881794552, |
|
"eval_loss": 0.10690969228744507, |
|
"eval_roc_auc": 0.8848957723927348, |
|
"eval_runtime": 1011.7793, |
|
"eval_samples_per_second": 2.833, |
|
"eval_steps_per_second": 0.089, |
|
"learning_rate": 0.001, |
|
"step": 1370 |
|
}, |
|
{ |
|
"epoch": 5.47, |
|
"learning_rate": 0.001, |
|
"loss": 0.1181, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.5212840195394278, |
|
"eval_f1_macro": 0.6806419028627124, |
|
"eval_f1_micro": 0.82229280970257, |
|
"eval_loss": 0.10277453809976578, |
|
"eval_roc_auc": 0.8815633385282128, |
|
"eval_runtime": 1010.7838, |
|
"eval_samples_per_second": 2.835, |
|
"eval_steps_per_second": 0.089, |
|
"learning_rate": 0.001, |
|
"step": 1644 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.5268667131891137, |
|
"eval_f1_macro": 0.6718237657927368, |
|
"eval_f1_micro": 0.8252749770852429, |
|
"eval_loss": 0.09917270392179489, |
|
"eval_roc_auc": 0.8857388585296528, |
|
"eval_runtime": 1012.914, |
|
"eval_samples_per_second": 2.829, |
|
"eval_steps_per_second": 0.089, |
|
"learning_rate": 0.001, |
|
"step": 1918 |
|
}, |
|
{ |
|
"epoch": 7.3, |
|
"learning_rate": 0.001, |
|
"loss": 0.1146, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.5216329378925332, |
|
"eval_f1_macro": 0.6810819602181296, |
|
"eval_f1_micro": 0.8223672809196465, |
|
"eval_loss": 0.10044297575950623, |
|
"eval_roc_auc": 0.8815306397502212, |
|
"eval_runtime": 1019.7188, |
|
"eval_samples_per_second": 2.811, |
|
"eval_steps_per_second": 0.088, |
|
"learning_rate": 0.001, |
|
"step": 2192 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.5230286113049546, |
|
"eval_f1_macro": 0.6845263067090662, |
|
"eval_f1_micro": 0.8302484858776249, |
|
"eval_loss": 0.10188507288694382, |
|
"eval_roc_auc": 0.8923158419824271, |
|
"eval_runtime": 1018.1313, |
|
"eval_samples_per_second": 2.815, |
|
"eval_steps_per_second": 0.088, |
|
"learning_rate": 0.001, |
|
"step": 2466 |
|
}, |
|
{ |
|
"epoch": 9.12, |
|
"learning_rate": 0.001, |
|
"loss": 0.1123, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.5279134682484299, |
|
"eval_f1_macro": 0.6775112836458498, |
|
"eval_f1_micro": 0.8181073809802316, |
|
"eval_loss": 0.10208763182163239, |
|
"eval_roc_auc": 0.8754023666126408, |
|
"eval_runtime": 1019.3535, |
|
"eval_samples_per_second": 2.812, |
|
"eval_steps_per_second": 0.088, |
|
"learning_rate": 0.001, |
|
"step": 2740 |
|
}, |
|
{ |
|
"epoch": 10.95, |
|
"learning_rate": 0.001, |
|
"loss": 0.11, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_accuracy": 0.5429169574319609, |
|
"eval_f1_macro": 0.6897433634697611, |
|
"eval_f1_micro": 0.8290004089501665, |
|
"eval_loss": 0.09596865624189377, |
|
"eval_roc_auc": 0.8815259520118676, |
|
"eval_runtime": 1024.4911, |
|
"eval_samples_per_second": 2.797, |
|
"eval_steps_per_second": 0.088, |
|
"learning_rate": 0.001, |
|
"step": 3014 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.537334263782275, |
|
"eval_f1_macro": 0.682540187560418, |
|
"eval_f1_micro": 0.8315506492020821, |
|
"eval_loss": 0.0966910719871521, |
|
"eval_roc_auc": 0.8896335469441946, |
|
"eval_runtime": 1022.2342, |
|
"eval_samples_per_second": 2.804, |
|
"eval_steps_per_second": 0.088, |
|
"learning_rate": 0.001, |
|
"step": 3288 |
|
}, |
|
{ |
|
"epoch": 12.77, |
|
"learning_rate": 0.001, |
|
"loss": 0.1098, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_accuracy": 0.5327983251919051, |
|
"eval_f1_macro": 0.6961160759945566, |
|
"eval_f1_micro": 0.8254151291512914, |
|
"eval_loss": 0.10087021440267563, |
|
"eval_roc_auc": 0.8837691574163886, |
|
"eval_runtime": 1018.0187, |
|
"eval_samples_per_second": 2.815, |
|
"eval_steps_per_second": 0.088, |
|
"learning_rate": 0.001, |
|
"step": 3562 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.5331472435450104, |
|
"eval_f1_macro": 0.7092151646695873, |
|
"eval_f1_micro": 0.8278460468351495, |
|
"eval_loss": 0.09922181814908981, |
|
"eval_roc_auc": 0.8833891332859308, |
|
"eval_runtime": 691.5955, |
|
"eval_samples_per_second": 4.144, |
|
"eval_steps_per_second": 0.13, |
|
"learning_rate": 0.001, |
|
"step": 3836 |
|
}, |
|
{ |
|
"epoch": 14.6, |
|
"learning_rate": 0.001, |
|
"loss": 0.1166, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_accuracy": 0.5334961618981159, |
|
"eval_f1_macro": 0.6923192725695103, |
|
"eval_f1_micro": 0.8229681773227064, |
|
"eval_loss": 0.09913520514965057, |
|
"eval_roc_auc": 0.8796342115933741, |
|
"eval_runtime": 682.6982, |
|
"eval_samples_per_second": 4.198, |
|
"eval_steps_per_second": 0.132, |
|
"learning_rate": 0.001, |
|
"step": 4110 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.5390788555478018, |
|
"eval_f1_macro": 0.7123997578521556, |
|
"eval_f1_micro": 0.8394566401100476, |
|
"eval_loss": 0.09176727384328842, |
|
"eval_roc_auc": 0.8932457594574617, |
|
"eval_runtime": 684.6393, |
|
"eval_samples_per_second": 4.186, |
|
"eval_steps_per_second": 0.131, |
|
"learning_rate": 0.0001, |
|
"step": 4384 |
|
}, |
|
{ |
|
"epoch": 16.42, |
|
"learning_rate": 0.0001, |
|
"loss": 0.1072, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_accuracy": 0.5471039776692254, |
|
"eval_f1_macro": 0.7179397631117517, |
|
"eval_f1_micro": 0.8438971798218237, |
|
"eval_loss": 0.09066394716501236, |
|
"eval_roc_auc": 0.898963054967641, |
|
"eval_runtime": 690.5455, |
|
"eval_samples_per_second": 4.15, |
|
"eval_steps_per_second": 0.13, |
|
"learning_rate": 0.0001, |
|
"step": 4658 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.5446615491974878, |
|
"eval_f1_macro": 0.7236976914894626, |
|
"eval_f1_micro": 0.8441121068554142, |
|
"eval_loss": 0.08911634236574173, |
|
"eval_roc_auc": 0.8971207593709188, |
|
"eval_runtime": 685.8059, |
|
"eval_samples_per_second": 4.179, |
|
"eval_steps_per_second": 0.131, |
|
"learning_rate": 0.0001, |
|
"step": 4932 |
|
}, |
|
{ |
|
"epoch": 18.25, |
|
"learning_rate": 0.0001, |
|
"loss": 0.1017, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"eval_accuracy": 0.5478018143754362, |
|
"eval_f1_macro": 0.727715453888227, |
|
"eval_f1_micro": 0.8479439674649797, |
|
"eval_loss": 0.08924739807844162, |
|
"eval_roc_auc": 0.9027101012597639, |
|
"eval_runtime": 697.8818, |
|
"eval_samples_per_second": 4.107, |
|
"eval_steps_per_second": 0.129, |
|
"learning_rate": 0.0001, |
|
"step": 5206 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"eval_accuracy": 0.5509420795533845, |
|
"eval_f1_macro": 0.7309632047081221, |
|
"eval_f1_micro": 0.848399617483265, |
|
"eval_loss": 0.08830931037664413, |
|
"eval_roc_auc": 0.9043454872974694, |
|
"eval_runtime": 680.9035, |
|
"eval_samples_per_second": 4.209, |
|
"eval_steps_per_second": 0.132, |
|
"learning_rate": 0.0001, |
|
"step": 5480 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"learning_rate": 0.0001, |
|
"step": 5480, |
|
"total_flos": 2.5999409342512156e+20, |
|
"train_loss": 0.03709733433967089, |
|
"train_runtime": 20246.0375, |
|
"train_samples_per_second": 8.66, |
|
"train_steps_per_second": 0.271 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 5480, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 20, |
|
"save_steps": 500, |
|
"total_flos": 2.5999409342512156e+20, |
|
"train_batch_size": 32, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|