|
{ |
|
"best_metric": 0.7985673319510695, |
|
"best_model_checkpoint": "InLegalBERT/checkpoint-3215", |
|
"epoch": 10.0, |
|
"eval_steps": 500, |
|
"global_step": 6430, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.78, |
|
"grad_norm": 6.225058078765869, |
|
"learning_rate": 4.742871954380509e-05, |
|
"loss": 1.065, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.7993803253292022, |
|
"eval_f1_macro": 0.6028764374506596, |
|
"eval_f1_micro": 0.7993803253292022, |
|
"eval_f1_weighted": 0.780434475079005, |
|
"eval_loss": 0.6395189166069031, |
|
"eval_macro_fpr": 0.01853296480715729, |
|
"eval_macro_sensitivity": 0.6307714529979154, |
|
"eval_macro_specificity": 0.9847175193097543, |
|
"eval_precision": 0.7817573111259606, |
|
"eval_precision_macro": 0.6194286814679265, |
|
"eval_recall": 0.7993803253292022, |
|
"eval_recall_macro": 0.6307714529979154, |
|
"eval_runtime": 29.4696, |
|
"eval_samples_per_second": 43.808, |
|
"eval_steps_per_second": 5.497, |
|
"eval_weighted_fpr": 0.017610661589719183, |
|
"eval_weighted_sensitivity": 0.7993803253292022, |
|
"eval_weighted_specificity": 0.9713824643171124, |
|
"step": 643 |
|
}, |
|
{ |
|
"epoch": 1.56, |
|
"grad_norm": 8.756747245788574, |
|
"learning_rate": 4.483670295489891e-05, |
|
"loss": 0.5866, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.8187451587916343, |
|
"eval_f1_macro": 0.7276084501581802, |
|
"eval_f1_micro": 0.8187451587916342, |
|
"eval_f1_weighted": 0.8152450677099363, |
|
"eval_loss": 0.6906521916389465, |
|
"eval_macro_fpr": 0.01610571025604924, |
|
"eval_macro_sensitivity": 0.7365748487318147, |
|
"eval_macro_specificity": 0.9863504064994671, |
|
"eval_precision": 0.8198905458065566, |
|
"eval_precision_macro": 0.7284980957351811, |
|
"eval_recall": 0.8187451587916343, |
|
"eval_recall_macro": 0.7365748487318147, |
|
"eval_runtime": 57.1075, |
|
"eval_samples_per_second": 22.606, |
|
"eval_steps_per_second": 2.837, |
|
"eval_weighted_fpr": 0.015566790846194785, |
|
"eval_weighted_sensitivity": 0.8187451587916343, |
|
"eval_weighted_specificity": 0.976510938700373, |
|
"step": 1286 |
|
}, |
|
{ |
|
"epoch": 2.33, |
|
"grad_norm": 8.241662979125977, |
|
"learning_rate": 4.224468636599274e-05, |
|
"loss": 0.4622, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.8179705654531371, |
|
"eval_f1_macro": 0.7282676953256336, |
|
"eval_f1_micro": 0.817970565453137, |
|
"eval_f1_weighted": 0.8150036178041801, |
|
"eval_loss": 0.8055915236473083, |
|
"eval_macro_fpr": 0.016168738933500626, |
|
"eval_macro_sensitivity": 0.737597935006265, |
|
"eval_macro_specificity": 0.9862931306285473, |
|
"eval_precision": 0.8137371356291269, |
|
"eval_precision_macro": 0.7226669118270935, |
|
"eval_recall": 0.8179705654531371, |
|
"eval_recall_macro": 0.737597935006265, |
|
"eval_runtime": 56.5581, |
|
"eval_samples_per_second": 22.826, |
|
"eval_steps_per_second": 2.864, |
|
"eval_weighted_fpr": 0.01564684732671949, |
|
"eval_weighted_sensitivity": 0.8179705654531371, |
|
"eval_weighted_specificity": 0.9764263939750704, |
|
"step": 1929 |
|
}, |
|
{ |
|
"epoch": 3.11, |
|
"grad_norm": 5.718050003051758, |
|
"learning_rate": 3.9652669777086575e-05, |
|
"loss": 0.346, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 3.89, |
|
"grad_norm": 28.698440551757812, |
|
"learning_rate": 3.7060653188180404e-05, |
|
"loss": 0.2398, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.8171959721146398, |
|
"eval_f1_macro": 0.7407336443567221, |
|
"eval_f1_micro": 0.8171959721146398, |
|
"eval_f1_weighted": 0.8160506897887181, |
|
"eval_loss": 0.9310164451599121, |
|
"eval_macro_fpr": 0.016134598971936556, |
|
"eval_macro_sensitivity": 0.7424877811517157, |
|
"eval_macro_specificity": 0.9862275325034797, |
|
"eval_precision": 0.8234764041889034, |
|
"eval_precision_macro": 0.7660803499273104, |
|
"eval_recall": 0.8171959721146398, |
|
"eval_recall_macro": 0.7424877811517157, |
|
"eval_runtime": 58.71, |
|
"eval_samples_per_second": 21.989, |
|
"eval_steps_per_second": 2.759, |
|
"eval_weighted_fpr": 0.015727042516326804, |
|
"eval_weighted_sensitivity": 0.8171959721146398, |
|
"eval_weighted_specificity": 0.9762170154375561, |
|
"step": 2572 |
|
}, |
|
{ |
|
"epoch": 4.67, |
|
"grad_norm": 8.432526588439941, |
|
"learning_rate": 3.447382063245205e-05, |
|
"loss": 0.1611, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.8303640588690937, |
|
"eval_f1_macro": 0.7985673319510695, |
|
"eval_f1_micro": 0.8303640588690937, |
|
"eval_f1_weighted": 0.8304130023375266, |
|
"eval_loss": 1.0763046741485596, |
|
"eval_macro_fpr": 0.014791024663918677, |
|
"eval_macro_sensitivity": 0.7918299512118986, |
|
"eval_macro_specificity": 0.9872504344639083, |
|
"eval_precision": 0.8362631231068205, |
|
"eval_precision_macro": 0.8174407838611729, |
|
"eval_recall": 0.8303640588690937, |
|
"eval_recall_macro": 0.7918299512118986, |
|
"eval_runtime": 71.8128, |
|
"eval_samples_per_second": 17.977, |
|
"eval_steps_per_second": 2.256, |
|
"eval_weighted_fpr": 0.014382347146516056, |
|
"eval_weighted_sensitivity": 0.8303640588690937, |
|
"eval_weighted_specificity": 0.978392458089531, |
|
"step": 3215 |
|
}, |
|
{ |
|
"epoch": 5.44, |
|
"grad_norm": 0.10359270125627518, |
|
"learning_rate": 3.1881804043545884e-05, |
|
"loss": 0.1055, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.82571649883811, |
|
"eval_f1_macro": 0.786305894104014, |
|
"eval_f1_micro": 0.82571649883811, |
|
"eval_f1_weighted": 0.824551084474796, |
|
"eval_loss": 1.1377127170562744, |
|
"eval_macro_fpr": 0.015367631824352218, |
|
"eval_macro_sensitivity": 0.7810139173900718, |
|
"eval_macro_specificity": 0.9868830526127108, |
|
"eval_precision": 0.8275394319817495, |
|
"eval_precision_macro": 0.8039087862954798, |
|
"eval_recall": 0.82571649883811, |
|
"eval_recall_macro": 0.7810139173900718, |
|
"eval_runtime": 64.6694, |
|
"eval_samples_per_second": 19.963, |
|
"eval_steps_per_second": 2.505, |
|
"eval_weighted_fpr": 0.01485246550927454, |
|
"eval_weighted_sensitivity": 0.82571649883811, |
|
"eval_weighted_specificity": 0.9775292903525532, |
|
"step": 3858 |
|
}, |
|
{ |
|
"epoch": 6.22, |
|
"grad_norm": 0.07212593406438828, |
|
"learning_rate": 2.928978745463971e-05, |
|
"loss": 0.1152, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"grad_norm": 0.15883693099021912, |
|
"learning_rate": 2.669777086573354e-05, |
|
"loss": 0.0463, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.8071262587141751, |
|
"eval_f1_macro": 0.7661147949067431, |
|
"eval_f1_micro": 0.8071262587141751, |
|
"eval_f1_weighted": 0.8078396785965507, |
|
"eval_loss": 1.321502447128296, |
|
"eval_macro_fpr": 0.017229971151571283, |
|
"eval_macro_sensitivity": 0.7688718615633127, |
|
"eval_macro_specificity": 0.9855509290260365, |
|
"eval_precision": 0.8111033305830805, |
|
"eval_precision_macro": 0.7691608912363075, |
|
"eval_recall": 0.8071262587141751, |
|
"eval_recall_macro": 0.7688718615633127, |
|
"eval_runtime": 77.5552, |
|
"eval_samples_per_second": 16.646, |
|
"eval_steps_per_second": 2.089, |
|
"eval_weighted_fpr": 0.016782368403316036, |
|
"eval_weighted_sensitivity": 0.8071262587141751, |
|
"eval_weighted_specificity": 0.9761376766763701, |
|
"step": 4501 |
|
}, |
|
{ |
|
"epoch": 7.78, |
|
"grad_norm": 0.003739361884072423, |
|
"learning_rate": 2.4105754276827372e-05, |
|
"loss": 0.031, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.820294345468629, |
|
"eval_f1_macro": 0.7689822136833361, |
|
"eval_f1_micro": 0.8202943454686291, |
|
"eval_f1_weighted": 0.8174949305492484, |
|
"eval_loss": 1.3483381271362305, |
|
"eval_macro_fpr": 0.01612405504801943, |
|
"eval_macro_sensitivity": 0.7726825622003671, |
|
"eval_macro_specificity": 0.9863605879945143, |
|
"eval_precision": 0.8169871783387602, |
|
"eval_precision_macro": 0.7773067113000881, |
|
"eval_recall": 0.820294345468629, |
|
"eval_recall_macro": 0.7726825622003671, |
|
"eval_runtime": 57.45, |
|
"eval_samples_per_second": 22.472, |
|
"eval_steps_per_second": 2.82, |
|
"eval_weighted_fpr": 0.015407092575375215, |
|
"eval_weighted_sensitivity": 0.820294345468629, |
|
"eval_weighted_specificity": 0.9751144744490858, |
|
"step": 5144 |
|
}, |
|
{ |
|
"epoch": 8.55, |
|
"grad_norm": 0.008842483162879944, |
|
"learning_rate": 2.151892172109902e-05, |
|
"loss": 0.0202, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.8280402788536019, |
|
"eval_f1_macro": 0.7753479744325771, |
|
"eval_f1_micro": 0.8280402788536019, |
|
"eval_f1_weighted": 0.825561089776041, |
|
"eval_loss": 1.3729532957077026, |
|
"eval_macro_fpr": 0.015177813835714501, |
|
"eval_macro_sensitivity": 0.7802568330601105, |
|
"eval_macro_specificity": 0.9870641550691818, |
|
"eval_precision": 0.8263269633953413, |
|
"eval_precision_macro": 0.7817620978620624, |
|
"eval_recall": 0.8280402788536019, |
|
"eval_recall_macro": 0.7802568330601105, |
|
"eval_runtime": 69.0336, |
|
"eval_samples_per_second": 18.701, |
|
"eval_steps_per_second": 2.347, |
|
"eval_weighted_fpr": 0.014616802739004477, |
|
"eval_weighted_sensitivity": 0.8280402788536019, |
|
"eval_weighted_specificity": 0.977922047184124, |
|
"step": 5787 |
|
}, |
|
{ |
|
"epoch": 9.33, |
|
"grad_norm": 0.0021986050996929407, |
|
"learning_rate": 1.8926905132192844e-05, |
|
"loss": 0.0133, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.8164213787761425, |
|
"eval_f1_macro": 0.7655491262772139, |
|
"eval_f1_micro": 0.8164213787761425, |
|
"eval_f1_weighted": 0.8134862058117613, |
|
"eval_loss": 1.5407140254974365, |
|
"eval_macro_fpr": 0.016532295658760295, |
|
"eval_macro_sensitivity": 0.7779131623925402, |
|
"eval_macro_specificity": 0.9861038702342115, |
|
"eval_precision": 0.8162862069364494, |
|
"eval_precision_macro": 0.7687513177071161, |
|
"eval_recall": 0.8164213787761425, |
|
"eval_recall_macro": 0.7779131623925402, |
|
"eval_runtime": 56.3906, |
|
"eval_samples_per_second": 22.894, |
|
"eval_steps_per_second": 2.873, |
|
"eval_weighted_fpr": 0.01580737677582872, |
|
"eval_weighted_sensitivity": 0.8164213787761425, |
|
"eval_weighted_specificity": 0.9751366747370328, |
|
"step": 6430 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 9645, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 15, |
|
"save_steps": 500, |
|
"total_flos": 1.352022394447872e+16, |
|
"train_batch_size": 8, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|