cite-text-analysis's picture
Upload folder using huggingface_hub
de2cd4d verified
raw
history blame
15.5 kB
{
"best_metric": 0.6528931773500395,
"best_model_checkpoint": "case-analysis-bert-base-uncased/checkpoint-2912",
"epoch": 16.0,
"eval_steps": 500,
"global_step": 3584,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_accuracy": 0.7861915367483296,
"eval_f1_macro": 0.5606209513919795,
"eval_f1_micro": 0.7861915367483298,
"eval_f1_weighted": 0.7596863457967054,
"eval_loss": 0.7283128499984741,
"eval_macro_fpr": 0.1142150114777894,
"eval_macro_sensitivity": 0.5571958513134984,
"eval_macro_specificity": 0.8974437739369907,
"eval_precision": 0.7487100743874461,
"eval_precision_macro": 0.584767668299669,
"eval_recall": 0.7861915367483296,
"eval_recall_macro": 0.5571958513134984,
"eval_runtime": 9.7006,
"eval_samples_per_second": 46.286,
"eval_steps_per_second": 5.876,
"eval_weighted_fpr": 0.08311688311688312,
"eval_weighted_sensitivity": 0.7861915367483296,
"eval_weighted_specificity": 0.8035835589996332,
"step": 224
},
{
"epoch": 2.0,
"eval_accuracy": 0.799554565701559,
"eval_f1_macro": 0.5913567856292314,
"eval_f1_micro": 0.799554565701559,
"eval_f1_weighted": 0.7794312570974847,
"eval_loss": 0.816001296043396,
"eval_macro_fpr": 0.09972743460349152,
"eval_macro_sensitivity": 0.6065148418089594,
"eval_macro_specificity": 0.910318444398066,
"eval_precision": 0.7603126937076594,
"eval_precision_macro": 0.5769576335614072,
"eval_recall": 0.799554565701559,
"eval_recall_macro": 0.6065148418089594,
"eval_runtime": 39.2963,
"eval_samples_per_second": 11.426,
"eval_steps_per_second": 1.451,
"eval_weighted_fpr": 0.07712082262210797,
"eval_weighted_sensitivity": 0.799554565701559,
"eval_weighted_specificity": 0.8417192118907049,
"step": 448
},
{
"epoch": 2.232142857142857,
"grad_norm": 10.089143753051758,
"learning_rate": 4.6302083333333335e-05,
"loss": 0.6512,
"step": 500
},
{
"epoch": 3.0,
"eval_accuracy": 0.7906458797327395,
"eval_f1_macro": 0.5840451851613266,
"eval_f1_micro": 0.7906458797327395,
"eval_f1_weighted": 0.771952482046502,
"eval_loss": 0.8587548136711121,
"eval_macro_fpr": 0.10045553338392213,
"eval_macro_sensitivity": 0.5989408048231577,
"eval_macro_specificity": 0.9104668136432382,
"eval_precision": 0.7597703984690992,
"eval_precision_macro": 0.5770004879708264,
"eval_recall": 0.7906458797327395,
"eval_recall_macro": 0.5989408048231577,
"eval_runtime": 36.0068,
"eval_samples_per_second": 12.47,
"eval_steps_per_second": 1.583,
"eval_weighted_fpr": 0.0811044003451251,
"eval_weighted_sensitivity": 0.7906458797327395,
"eval_weighted_specificity": 0.8512213748402133,
"step": 672
},
{
"epoch": 4.0,
"eval_accuracy": 0.7817371937639198,
"eval_f1_macro": 0.6299493260025852,
"eval_f1_micro": 0.7817371937639198,
"eval_f1_weighted": 0.7804892003577936,
"eval_loss": 1.0820659399032593,
"eval_macro_fpr": 0.09955325556886446,
"eval_macro_sensitivity": 0.6429130620307091,
"eval_macro_specificity": 0.9124114631310538,
"eval_precision": 0.7819490155060428,
"eval_precision_macro": 0.6213857962213225,
"eval_recall": 0.7817371937639198,
"eval_recall_macro": 0.6429130620307091,
"eval_runtime": 32.7545,
"eval_samples_per_second": 13.708,
"eval_steps_per_second": 1.74,
"eval_weighted_fpr": 0.08514335360556038,
"eval_weighted_sensitivity": 0.7817371937639198,
"eval_weighted_specificity": 0.8679086587602954,
"step": 896
},
{
"epoch": 4.464285714285714,
"grad_norm": 0.3640059232711792,
"learning_rate": 4.258184523809524e-05,
"loss": 0.3466,
"step": 1000
},
{
"epoch": 5.0,
"eval_accuracy": 0.8084632516703786,
"eval_f1_macro": 0.619470115868435,
"eval_f1_micro": 0.8084632516703786,
"eval_f1_weighted": 0.7927603955642376,
"eval_loss": 1.0612273216247559,
"eval_macro_fpr": 0.09480480467421379,
"eval_macro_sensitivity": 0.6262706556824204,
"eval_macro_specificity": 0.9138770264258179,
"eval_precision": 0.7999415514782998,
"eval_precision_macro": 0.7128659611992946,
"eval_recall": 0.8084632516703786,
"eval_recall_macro": 0.6262706556824204,
"eval_runtime": 34.7281,
"eval_samples_per_second": 12.929,
"eval_steps_per_second": 1.641,
"eval_weighted_fpr": 0.07319148936170213,
"eval_weighted_sensitivity": 0.8084632516703786,
"eval_weighted_specificity": 0.847044854032893,
"step": 1120
},
{
"epoch": 6.0,
"eval_accuracy": 0.7928730512249443,
"eval_f1_macro": 0.6273476992208005,
"eval_f1_micro": 0.7928730512249443,
"eval_f1_weighted": 0.7896519709076985,
"eval_loss": 1.2559400796890259,
"eval_macro_fpr": 0.09508832825281638,
"eval_macro_sensitivity": 0.6361528640940406,
"eval_macro_specificity": 0.9161350599199577,
"eval_precision": 0.7876580739535682,
"eval_precision_macro": 0.6205792331164909,
"eval_recall": 0.7928730512249443,
"eval_recall_macro": 0.6361528640940406,
"eval_runtime": 38.0322,
"eval_samples_per_second": 11.806,
"eval_steps_per_second": 1.499,
"eval_weighted_fpr": 0.08010335917312661,
"eval_weighted_sensitivity": 0.7928730512249443,
"eval_weighted_specificity": 0.8716671884548861,
"step": 1344
},
{
"epoch": 6.696428571428571,
"grad_norm": 0.05509716644883156,
"learning_rate": 3.8869047619047625e-05,
"loss": 0.1715,
"step": 1500
},
{
"epoch": 7.0,
"eval_accuracy": 0.7928730512249443,
"eval_f1_macro": 0.623650691198171,
"eval_f1_micro": 0.7928730512249443,
"eval_f1_weighted": 0.7892792124992886,
"eval_loss": 1.3700848817825317,
"eval_macro_fpr": 0.09905686727492574,
"eval_macro_sensitivity": 0.6178801987625516,
"eval_macro_specificity": 0.9121558558439,
"eval_precision": 0.7889374292245007,
"eval_precision_macro": 0.6344745555953697,
"eval_recall": 0.7928730512249443,
"eval_recall_macro": 0.6178801987625516,
"eval_runtime": 38.6276,
"eval_samples_per_second": 11.624,
"eval_steps_per_second": 1.476,
"eval_weighted_fpr": 0.08010335917312661,
"eval_weighted_sensitivity": 0.7928730512249443,
"eval_weighted_specificity": 0.8557503721506562,
"step": 1568
},
{
"epoch": 8.0,
"eval_accuracy": 0.8106904231625836,
"eval_f1_macro": 0.6464437798898967,
"eval_f1_micro": 0.8106904231625834,
"eval_f1_weighted": 0.8064189899649085,
"eval_loss": 1.4005118608474731,
"eval_macro_fpr": 0.0921754708102674,
"eval_macro_sensitivity": 0.6370398282162988,
"eval_macro_specificity": 0.9178527413440807,
"eval_precision": 0.8034816923327339,
"eval_precision_macro": 0.6578060011883542,
"eval_recall": 0.8106904231625836,
"eval_recall_macro": 0.6370398282162988,
"eval_runtime": 35.6698,
"eval_samples_per_second": 12.588,
"eval_steps_per_second": 1.598,
"eval_weighted_fpr": 0.07221750212404418,
"eval_weighted_sensitivity": 0.8106904231625836,
"eval_weighted_specificity": 0.860720542213739,
"step": 1792
},
{
"epoch": 8.928571428571429,
"grad_norm": 0.005432427860796452,
"learning_rate": 3.514880952380952e-05,
"loss": 0.0636,
"step": 2000
},
{
"epoch": 9.0,
"eval_accuracy": 0.8017817371937639,
"eval_f1_macro": 0.6263115408785538,
"eval_f1_micro": 0.8017817371937639,
"eval_f1_weighted": 0.7896240754364972,
"eval_loss": 1.473650336265564,
"eval_macro_fpr": 0.10259295399855256,
"eval_macro_sensitivity": 0.6149052987288282,
"eval_macro_specificity": 0.9072211454744684,
"eval_precision": 0.788055851480147,
"eval_precision_macro": 0.6582800876341822,
"eval_recall": 0.8017817371937639,
"eval_recall_macro": 0.6149052987288282,
"eval_runtime": 34.11,
"eval_samples_per_second": 13.163,
"eval_steps_per_second": 1.671,
"eval_weighted_fpr": 0.07613344739093242,
"eval_weighted_sensitivity": 0.8017817371937639,
"eval_weighted_specificity": 0.8271028447041098,
"step": 2016
},
{
"epoch": 10.0,
"eval_accuracy": 0.7884187082405345,
"eval_f1_macro": 0.6331690370327951,
"eval_f1_micro": 0.7884187082405345,
"eval_f1_weighted": 0.7909426185077314,
"eval_loss": 1.7568920850753784,
"eval_macro_fpr": 0.09595908153187079,
"eval_macro_sensitivity": 0.6428373560726501,
"eval_macro_specificity": 0.91584307980206,
"eval_precision": 0.7961677526311304,
"eval_precision_macro": 0.6275270485563467,
"eval_recall": 0.7884187082405345,
"eval_recall_macro": 0.6428373560726501,
"eval_runtime": 42.7436,
"eval_samples_per_second": 10.504,
"eval_steps_per_second": 1.334,
"eval_weighted_fpr": 0.08210890233362143,
"eval_weighted_sensitivity": 0.7884187082405345,
"eval_weighted_specificity": 0.8749536109677053,
"step": 2240
},
{
"epoch": 11.0,
"eval_accuracy": 0.7906458797327395,
"eval_f1_macro": 0.6101321036639349,
"eval_f1_micro": 0.7906458797327395,
"eval_f1_weighted": 0.7845175231806186,
"eval_loss": 1.7140623331069946,
"eval_macro_fpr": 0.1035198920558284,
"eval_macro_sensitivity": 0.6082973730032554,
"eval_macro_specificity": 0.9082706582818566,
"eval_precision": 0.7824040660397941,
"eval_precision_macro": 0.6165716760629552,
"eval_recall": 0.7906458797327395,
"eval_recall_macro": 0.6082973730032554,
"eval_runtime": 37.438,
"eval_samples_per_second": 11.993,
"eval_steps_per_second": 1.523,
"eval_weighted_fpr": 0.0811044003451251,
"eval_weighted_sensitivity": 0.7906458797327395,
"eval_weighted_specificity": 0.842436753394687,
"step": 2464
},
{
"epoch": 11.160714285714286,
"grad_norm": 0.0017704592319205403,
"learning_rate": 3.142857142857143e-05,
"loss": 0.0159,
"step": 2500
},
{
"epoch": 12.0,
"eval_accuracy": 0.7951002227171492,
"eval_f1_macro": 0.6372747490500124,
"eval_f1_micro": 0.7951002227171492,
"eval_f1_weighted": 0.7916670590899947,
"eval_loss": 1.714410662651062,
"eval_macro_fpr": 0.09685632952239012,
"eval_macro_sensitivity": 0.6412759206876854,
"eval_macro_specificity": 0.9140152346037369,
"eval_precision": 0.7913770752065027,
"eval_precision_macro": 0.6393393002684569,
"eval_recall": 0.7951002227171492,
"eval_recall_macro": 0.6412759206876854,
"eval_runtime": 34.7517,
"eval_samples_per_second": 12.92,
"eval_steps_per_second": 1.64,
"eval_weighted_fpr": 0.07910576096302666,
"eval_weighted_sensitivity": 0.7951002227171492,
"eval_weighted_specificity": 0.8609607156977986,
"step": 2688
},
{
"epoch": 13.0,
"eval_accuracy": 0.799554565701559,
"eval_f1_macro": 0.6528931773500395,
"eval_f1_micro": 0.799554565701559,
"eval_f1_weighted": 0.798188843376642,
"eval_loss": 1.7243342399597168,
"eval_macro_fpr": 0.09420589017594908,
"eval_macro_sensitivity": 0.6525655716832187,
"eval_macro_specificity": 0.9158469519501292,
"eval_precision": 0.7968817042081011,
"eval_precision_macro": 0.6534600751896625,
"eval_recall": 0.799554565701559,
"eval_recall_macro": 0.6525655716832187,
"eval_runtime": 35.7184,
"eval_samples_per_second": 12.571,
"eval_steps_per_second": 1.596,
"eval_weighted_fpr": 0.07712082262210797,
"eval_weighted_sensitivity": 0.799554565701559,
"eval_weighted_specificity": 0.8638332420989578,
"step": 2912
},
{
"epoch": 13.392857142857142,
"grad_norm": 0.0014869036385789514,
"learning_rate": 2.7708333333333337e-05,
"loss": 0.0043,
"step": 3000
},
{
"epoch": 14.0,
"eval_accuracy": 0.7973273942093542,
"eval_f1_macro": 0.6334742709127367,
"eval_f1_micro": 0.7973273942093542,
"eval_f1_weighted": 0.7911632678226206,
"eval_loss": 1.8550578355789185,
"eval_macro_fpr": 0.10413380473626513,
"eval_macro_sensitivity": 0.6189159939159938,
"eval_macro_specificity": 0.9071845213512801,
"eval_precision": 0.7948329741898784,
"eval_precision_macro": 0.6575689935064936,
"eval_recall": 0.7973273942093542,
"eval_recall_macro": 0.6189159939159938,
"eval_runtime": 37.3063,
"eval_samples_per_second": 12.035,
"eval_steps_per_second": 1.528,
"eval_weighted_fpr": 0.07811158798283262,
"eval_weighted_sensitivity": 0.7973273942093542,
"eval_weighted_specificity": 0.8314106911957669,
"step": 3136
},
{
"epoch": 15.0,
"eval_accuracy": 0.7928730512249443,
"eval_f1_macro": 0.6104462132353847,
"eval_f1_micro": 0.7928730512249443,
"eval_f1_weighted": 0.7860581271310794,
"eval_loss": 1.8840636014938354,
"eval_macro_fpr": 0.10077968846838097,
"eval_macro_sensitivity": 0.616244778009484,
"eval_macro_specificity": 0.9109811795750432,
"eval_precision": 0.7868702804113628,
"eval_precision_macro": 0.6153843681293956,
"eval_recall": 0.7928730512249443,
"eval_recall_macro": 0.616244778009484,
"eval_runtime": 37.5124,
"eval_samples_per_second": 11.969,
"eval_steps_per_second": 1.519,
"eval_weighted_fpr": 0.08010335917312661,
"eval_weighted_sensitivity": 0.7928730512249443,
"eval_weighted_specificity": 0.8510516670752286,
"step": 3360
},
{
"epoch": 15.625,
"grad_norm": 0.0012946381466463208,
"learning_rate": 2.3995535714285717e-05,
"loss": 0.0029,
"step": 3500
},
{
"epoch": 16.0,
"eval_accuracy": 0.755011135857461,
"eval_f1_macro": 0.6015295009953532,
"eval_f1_micro": 0.755011135857461,
"eval_f1_weighted": 0.7660603471590911,
"eval_loss": 2.0852534770965576,
"eval_macro_fpr": 0.1100287944234454,
"eval_macro_sensitivity": 0.6118684574566927,
"eval_macro_specificity": 0.9061965043355904,
"eval_precision": 0.7837194577525304,
"eval_precision_macro": 0.6010349959287868,
"eval_recall": 0.755011135857461,
"eval_recall_macro": 0.6118684574566927,
"eval_runtime": 40.7827,
"eval_samples_per_second": 11.01,
"eval_steps_per_second": 1.398,
"eval_weighted_fpr": 0.09760425909494233,
"eval_weighted_sensitivity": 0.755011135857461,
"eval_weighted_specificity": 0.8697748814849007,
"step": 3584
}
],
"logging_steps": 500,
"max_steps": 6720,
"num_input_tokens_seen": 0,
"num_train_epochs": 30,
"save_steps": 500,
"total_flos": 7544055646912512.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}