longformer-simple / checkpoint-1134 /trainer_state.json
Theoreticallyhugo's picture
Training in progress, epoch 14, checkpoint
e491f73 verified
raw
history blame
19.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 14.0,
"eval_steps": 500,
"global_step": 1134,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_Claim": {
"f1-score": 0.4888068880688806,
"precision": 0.6228840125391849,
"recall": 0.40222672064777326,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.7325278370054489,
"precision": 0.7604525332021643,
"recall": 0.706581352833638,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.8634848138844484,
"precision": 0.827613377692173,
"recall": 0.902606702950444,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8792819197655248,
"precision": 0.8543458914268763,
"recall": 0.9057173407132524,
"support": 15899.0
},
"eval_accuracy": 0.8174925373134329,
"eval_loss": 0.49806877970695496,
"eval_macro avg": {
"f1-score": 0.7410253646810757,
"precision": 0.7663239537150996,
"recall": 0.7292830292862769,
"support": 33500.0
},
"eval_runtime": 4.9194,
"eval_samples_per_second": 16.465,
"eval_steps_per_second": 2.236,
"eval_weighted avg": {
"f1-score": 0.8071778099281224,
"precision": 0.80572410140167,
"recall": 0.8174925373134329,
"support": 33500.0
},
"step": 81
},
{
"epoch": 2.0,
"eval_Claim": {
"f1-score": 0.5993816254416962,
"precision": 0.6593780369290574,
"recall": 0.5493927125506073,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8017817371937639,
"precision": 0.7197382769901853,
"recall": 0.9049360146252285,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9058997050147494,
"precision": 0.9337184554575859,
"recall": 0.8796906330564308,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8854737486606459,
"precision": 0.862579028987236,
"recall": 0.9096169570413234,
"support": 15899.0
},
"eval_accuracy": 0.8468358208955223,
"eval_loss": 0.4562840759754181,
"eval_macro avg": {
"f1-score": 0.7981342040777137,
"precision": 0.7938534495910161,
"recall": 0.8109090793183975,
"support": 33500.0
},
"eval_runtime": 4.904,
"eval_samples_per_second": 16.517,
"eval_steps_per_second": 2.243,
"eval_weighted avg": {
"f1-score": 0.8442053257085675,
"precision": 0.8455251408465503,
"recall": 0.8468358208955223,
"support": 33500.0
},
"step": 162
},
{
"epoch": 3.0,
"eval_Claim": {
"f1-score": 0.6235462150581514,
"precision": 0.6285479226655697,
"recall": 0.6186234817813765,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8338848444738584,
"precision": 0.8059701492537313,
"recall": 0.8638025594149908,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9066250826953974,
"precision": 0.8974646833193002,
"recall": 0.9159744103886184,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8827730692315017,
"precision": 0.8911176621379133,
"recall": 0.8745833071262343,
"support": 15899.0
},
"eval_accuracy": 0.8490746268656716,
"eval_loss": 0.43820467591285706,
"eval_macro avg": {
"f1-score": 0.8117073028647273,
"precision": 0.8057751043441286,
"recall": 0.818245939677805,
"support": 33500.0
},
"eval_runtime": 4.9179,
"eval_samples_per_second": 16.47,
"eval_steps_per_second": 2.237,
"eval_weighted avg": {
"f1-score": 0.8488105033097496,
"precision": 0.8488213959184715,
"recall": 0.8490746268656716,
"support": 33500.0
},
"step": 243
},
{
"epoch": 4.0,
"eval_Claim": {
"f1-score": 0.6415287144900879,
"precision": 0.6477507222451506,
"recall": 0.6354251012145749,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8005126014523707,
"precision": 0.7514033680834001,
"recall": 0.856489945155393,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9065169636319258,
"precision": 0.9273871354374751,
"recall": 0.8865654540246348,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8932505382719131,
"precision": 0.8863636363636364,
"recall": 0.9002452984464432,
"support": 15899.0
},
"eval_accuracy": 0.8540597014925373,
"eval_loss": 0.4749625325202942,
"eval_macro avg": {
"f1-score": 0.8104522044615743,
"precision": 0.8032262155324155,
"recall": 0.8196814497102615,
"support": 33500.0
},
"eval_runtime": 4.8972,
"eval_samples_per_second": 16.54,
"eval_steps_per_second": 2.246,
"eval_weighted avg": {
"f1-score": 0.8542213698406307,
"precision": 0.8551874645146165,
"recall": 0.8540597014925373,
"support": 33500.0
},
"step": 324
},
{
"epoch": 5.0,
"eval_Claim": {
"f1-score": 0.6099687080911935,
"precision": 0.6808882235528942,
"recall": 0.5524291497975709,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8315018315018317,
"precision": 0.8330275229357799,
"recall": 0.829981718464351,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9103454839630722,
"precision": 0.9074091642159188,
"recall": 0.9133008689009835,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8913988368533824,
"precision": 0.8682249120505635,
"recall": 0.915843763758727,
"support": 15899.0
},
"eval_accuracy": 0.8558507462686568,
"eval_loss": 0.5162495374679565,
"eval_macro avg": {
"f1-score": 0.8108037151023699,
"precision": 0.8223874556887891,
"recall": 0.8028888752304081,
"support": 33500.0
},
"eval_runtime": 4.9413,
"eval_samples_per_second": 16.392,
"eval_steps_per_second": 2.226,
"eval_weighted avg": {
"f1-score": 0.8519095458499607,
"precision": 0.8505509283002989,
"recall": 0.8558507462686568,
"support": 33500.0
},
"step": 405
},
{
"epoch": 6.0,
"eval_Claim": {
"f1-score": 0.646279703934554,
"precision": 0.6227477477477478,
"recall": 0.6716599190283401,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8414067684140676,
"precision": 0.8152593227603944,
"recall": 0.8692870201096892,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9112506693929215,
"precision": 0.9295788637266588,
"recall": 0.8936312422419556,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8861383012314494,
"precision": 0.8897343224906474,
"recall": 0.8825712308950249,
"support": 15899.0
},
"eval_accuracy": 0.8540597014925373,
"eval_loss": 0.552895724773407,
"eval_macro avg": {
"f1-score": 0.8212688607432481,
"precision": 0.8143300641813621,
"recall": 0.8292873530687525,
"support": 33500.0
},
"eval_runtime": 4.9336,
"eval_samples_per_second": 16.418,
"eval_steps_per_second": 2.23,
"eval_weighted avg": {
"f1-score": 0.8556973390614198,
"precision": 0.8579560210496036,
"recall": 0.8540597014925373,
"support": 33500.0
},
"step": 486
},
{
"epoch": 6.17,
"grad_norm": 9.790245056152344,
"learning_rate": 1.7530864197530865e-05,
"loss": 0.3328,
"step": 500
},
{
"epoch": 7.0,
"eval_Claim": {
"f1-score": 0.6202600808654791,
"precision": 0.6739491807171694,
"recall": 0.5744939271255061,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8670041417395307,
"precision": 0.8730305838739574,
"recall": 0.8610603290676416,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9109892729439809,
"precision": 0.9097314797181489,
"recall": 0.9122505490308412,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8872356123954747,
"precision": 0.8677611401767996,
"recall": 0.9076042518397384,
"support": 15899.0
},
"eval_accuracy": 0.8568955223880597,
"eval_loss": 0.5894299149513245,
"eval_macro avg": {
"f1-score": 0.8213722769861164,
"precision": 0.8311180961215189,
"recall": 0.8138522642659318,
"support": 33500.0
},
"eval_runtime": 4.9037,
"eval_samples_per_second": 16.518,
"eval_steps_per_second": 2.243,
"eval_weighted avg": {
"f1-score": 0.8539713289140156,
"precision": 0.8526463291050193,
"recall": 0.8568955223880597,
"support": 33500.0
},
"step": 567
},
{
"epoch": 8.0,
"eval_Claim": {
"f1-score": 0.6441500279902966,
"precision": 0.5974385600553825,
"recall": 0.6987854251012146,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.834510811389424,
"precision": 0.7849375755134917,
"recall": 0.8907678244972578,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9059161060546101,
"precision": 0.9398542543364466,
"recall": 0.874343550081161,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8801401497053671,
"precision": 0.8915849251419721,
"recall": 0.8689854707843261,
"support": 15899.0
},
"eval_accuracy": 0.8469850746268657,
"eval_loss": 0.5878348350524902,
"eval_macro avg": {
"f1-score": 0.8161792737849245,
"precision": 0.8034538287618231,
"recall": 0.8332205676159898,
"support": 33500.0
},
"eval_runtime": 4.9086,
"eval_samples_per_second": 16.502,
"eval_steps_per_second": 2.241,
"eval_weighted avg": {
"f1-score": 0.8504184600736624,
"precision": 0.8563340964894008,
"recall": 0.8469850746268657,
"support": 33500.0
},
"step": 648
},
{
"epoch": 9.0,
"eval_Claim": {
"f1-score": 0.6323094193285029,
"precision": 0.6375797489195307,
"recall": 0.6271255060728745,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8525914292124747,
"precision": 0.8373732921992068,
"recall": 0.8683729433272395,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9054132151801194,
"precision": 0.9134194927606647,
"recall": 0.8975460708488494,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8888055034396498,
"precision": 0.883775884584292,
"recall": 0.8938926976539405,
"support": 15899.0
},
"eval_accuracy": 0.8540298507462687,
"eval_loss": 0.7043101191520691,
"eval_macro avg": {
"f1-score": 0.8197798917901866,
"precision": 0.8180371046159236,
"recall": 0.821734304475726,
"support": 33500.0
},
"eval_runtime": 4.935,
"eval_samples_per_second": 16.414,
"eval_steps_per_second": 2.229,
"eval_weighted avg": {
"f1-score": 0.853808653145316,
"precision": 0.8537078167069386,
"recall": 0.8540298507462687,
"support": 33500.0
},
"step": 729
},
{
"epoch": 10.0,
"eval_Claim": {
"f1-score": 0.6347110795166041,
"precision": 0.6368453230079478,
"recall": 0.6325910931174089,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8442090055381651,
"precision": 0.8921119592875318,
"recall": 0.8011882998171846,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9106677408653618,
"precision": 0.9040270982310877,
"recall": 0.9174066647569942,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8841656478259506,
"precision": 0.881375,
"recall": 0.8869740235234921,
"support": 15899.0
},
"eval_accuracy": 0.8533731343283583,
"eval_loss": 0.7164433002471924,
"eval_macro avg": {
"f1-score": 0.8184383684365204,
"precision": 0.8285898451316418,
"recall": 0.8095400203037699,
"support": 33500.0
},
"eval_runtime": 4.9093,
"eval_samples_per_second": 16.499,
"eval_steps_per_second": 2.241,
"eval_weighted avg": {
"f1-score": 0.8530559678148433,
"precision": 0.8530989190255094,
"recall": 0.8533731343283583,
"support": 33500.0
},
"step": 810
},
{
"epoch": 11.0,
"eval_Claim": {
"f1-score": 0.645658668208538,
"precision": 0.6161486113665624,
"recall": 0.6781376518218624,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8331177231565331,
"precision": 0.7885714285714286,
"recall": 0.8829981718464351,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9059711059612445,
"precision": 0.9366843393148451,
"recall": 0.8772080588179128,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8902977542265961,
"precision": 0.8929452704840241,
"recall": 0.8876658909365369,
"support": 15899.0
},
"eval_accuracy": 0.8531940298507462,
"eval_loss": 0.7852362394332886,
"eval_macro avg": {
"f1-score": 0.818761312888228,
"precision": 0.8085874124342151,
"recall": 0.8315024433556868,
"support": 33500.0
},
"eval_runtime": 4.89,
"eval_samples_per_second": 16.565,
"eval_steps_per_second": 2.25,
"eval_weighted avg": {
"f1-score": 0.8553879040715653,
"precision": 0.8589850855801485,
"recall": 0.8531940298507462,
"support": 33500.0
},
"step": 891
},
{
"epoch": 12.0,
"eval_Claim": {
"f1-score": 0.6329215627196947,
"precision": 0.6280645804265498,
"recall": 0.6378542510121458,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8528368794326241,
"precision": 0.8278829604130808,
"recall": 0.8793418647166362,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9126755145220234,
"precision": 0.9193064031773709,
"recall": 0.9061395970591044,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.8876634630534112,
"precision": 0.8894291487749432,
"recall": 0.88590477388515,
"support": 15899.0
},
"eval_accuracy": 0.8552238805970149,
"eval_loss": 0.8324545621871948,
"eval_macro avg": {
"f1-score": 0.8215243549319383,
"precision": 0.8161707731979861,
"recall": 0.827310121668259,
"support": 33500.0
},
"eval_runtime": 4.8939,
"eval_samples_per_second": 16.551,
"eval_steps_per_second": 2.248,
"eval_weighted avg": {
"f1-score": 0.8556433335734092,
"precision": 0.8562082669116539,
"recall": 0.8552238805970149,
"support": 33500.0
},
"step": 972
},
{
"epoch": 12.35,
"grad_norm": 2.6090729236602783,
"learning_rate": 1.506172839506173e-05,
"loss": 0.0589,
"step": 1000
},
{
"epoch": 13.0,
"eval_Claim": {
"f1-score": 0.6431152147731835,
"precision": 0.6377388535031847,
"recall": 0.648582995951417,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8326953388876379,
"precision": 0.8206835330670218,
"recall": 0.8450639853747715,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.904438383337466,
"precision": 0.9408790755262072,
"recall": 0.8707151723479424,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.892013567684243,
"precision": 0.8749621922448733,
"recall": 0.9097427511164224,
"support": 15899.0
},
"eval_accuracy": 0.8548059701492537,
"eval_loss": 0.8543176651000977,
"eval_macro avg": {
"f1-score": 0.8180656261706326,
"precision": 0.8185659135853218,
"recall": 0.8185262261976383,
"support": 33500.0
},
"eval_runtime": 4.9338,
"eval_samples_per_second": 16.417,
"eval_steps_per_second": 2.23,
"eval_weighted avg": {
"f1-score": 0.8553204019036041,
"precision": 0.8570428644520474,
"recall": 0.8548059701492537,
"support": 33500.0
},
"step": 1053
},
{
"epoch": 14.0,
"eval_Claim": {
"f1-score": 0.6541843971631205,
"precision": 0.6138420585625555,
"recall": 0.7002024291497976,
"support": 4940.0
},
"eval_MajorClaim": {
"f1-score": 0.8336528221512247,
"precision": 0.7806142800159553,
"recall": 0.8944241316270567,
"support": 2188.0
},
"eval_O": {
"f1-score": 0.9079845651528643,
"precision": 0.9421004003695719,
"recall": 0.8762532225723288,
"support": 10473.0
},
"eval_Premise": {
"f1-score": 0.889389516436096,
"precision": 0.8974194787731319,
"recall": 0.8815019812566828,
"support": 15899.0
},
"eval_accuracy": 0.8539701492537314,
"eval_loss": 0.9095103144645691,
"eval_macro avg": {
"f1-score": 0.8213028252258264,
"precision": 0.8084940544303038,
"recall": 0.8380954411514664,
"support": 33500.0
},
"eval_runtime": 4.8917,
"eval_samples_per_second": 16.559,
"eval_steps_per_second": 2.249,
"eval_weighted avg": {
"f1-score": 0.8568784946124219,
"precision": 0.8619418985092084,
"recall": 0.8539701492537314,
"support": 33500.0
},
"step": 1134
}
],
"logging_steps": 500,
"max_steps": 4050,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"total_flos": 2006790574143600.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}