longformer-simple / checkpoint-1296 /trainer_state.json
Theoreticallyhugo's picture
Training in progress, epoch 16, checkpoint
8a1ff73 verified
raw
history blame
21.8 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 16.0,
"eval_steps": 500,
"global_step": 1296,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_Claim": {
"f1-score": 0.44734338904742493,
"precision": 0.5477537437603993,
"recall": 0.3780431786862655,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.6504965622612681,
"precision": 0.5898856944925528,
"recall": 0.7249893571732652,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.8859784283513097,
"precision": 0.8689081979599547,
"recall": 0.9037328094302554,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8720120805863504,
"precision": 0.859258183929738,
"recall": 0.8851502916105877,
"support": 13374.0
},
"eval_accuracy": 0.8059953068711373,
"eval_loss": 0.47828415036201477,
"eval_macro avg": {
"f1-score": 0.7139576150615883,
"precision": 0.7164514550356611,
"recall": 0.7229789092250934,
"support": 30257.0
},
"eval_runtime": 4.8422,
"eval_samples_per_second": 16.728,
"eval_steps_per_second": 2.272,
"eval_weighted avg": {
"f1-score": 0.7984036588902531,
"precision": 0.7967665566250599,
"recall": 0.8059953068711373,
"support": 30257.0
},
"step": 81
},
{
"epoch": 2.0,
"eval_Claim": {
"f1-score": 0.5787561000934482,
"precision": 0.5281409891984082,
"recall": 0.6401010564997703,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.7702650787552824,
"precision": 0.7017850892544627,
"recall": 0.8535547041294168,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9095337508698677,
"precision": 0.9206077681626081,
"recall": 0.8987229862475442,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8736648538675224,
"precision": 0.9162905211325401,
"recall": 0.8348287722446538,
"support": 13374.0
},
"eval_accuracy": 0.8297584030141786,
"eval_loss": 0.4241589605808258,
"eval_macro avg": {
"f1-score": 0.7830549458965301,
"precision": 0.7667060919370048,
"recall": 0.8068018797803463,
"support": 30257.0
},
"eval_runtime": 4.8636,
"eval_samples_per_second": 16.654,
"eval_steps_per_second": 2.262,
"eval_weighted avg": {
"f1-score": 0.8352680063880269,
"precision": 0.8452350051608073,
"recall": 0.8297584030141786,
"support": 30257.0
},
"step": 162
},
{
"epoch": 3.0,
"eval_Claim": {
"f1-score": 0.5728465083303793,
"precision": 0.5899245558530056,
"recall": 0.5567294441892513,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.7862876979822088,
"precision": 0.8017699115044248,
"recall": 0.7713920817369093,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.914532821534425,
"precision": 0.9048168445341794,
"recall": 0.9244597249508841,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8880533114924986,
"precision": 0.8843330614665975,
"recall": 0.8918049947659639,
"support": 13374.0
},
"eval_accuracy": 0.8452258981392735,
"eval_loss": 0.41067537665367126,
"eval_macro avg": {
"f1-score": 0.7904300848348779,
"precision": 0.7952110933395518,
"recall": 0.7860965614107521,
"support": 30257.0
},
"eval_runtime": 4.8684,
"eval_samples_per_second": 16.638,
"eval_steps_per_second": 2.259,
"eval_weighted avg": {
"f1-score": 0.8437033615676307,
"precision": 0.8424495118392472,
"recall": 0.8452258981392735,
"support": 30257.0
},
"step": 243
},
{
"epoch": 4.0,
"eval_Claim": {
"f1-score": 0.5865384615384616,
"precision": 0.5714596949891068,
"recall": 0.6024345429490124,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.7783630470016208,
"precision": 0.7425589485890993,
"recall": 0.8177948063005535,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9108591535529275,
"precision": 0.9063411787589963,
"recall": 0.9154223968565816,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8868256151612411,
"precision": 0.9067823097358962,
"recall": 0.8677284282937042,
"support": 13374.0
},
"eval_accuracy": 0.8417225765938461,
"eval_loss": 0.4545813202857971,
"eval_macro avg": {
"f1-score": 0.7906465693135627,
"precision": 0.7817855330182747,
"recall": 0.8008450435999629,
"support": 30257.0
},
"eval_runtime": 4.9112,
"eval_samples_per_second": 16.493,
"eval_steps_per_second": 2.24,
"eval_weighted avg": {
"f1-score": 0.8432797441676475,
"precision": 0.8456313015960876,
"recall": 0.8417225765938461,
"support": 30257.0
},
"step": 324
},
{
"epoch": 5.0,
"eval_Claim": {
"f1-score": 0.597399608790703,
"precision": 0.5985704403965875,
"recall": 0.5962333486449242,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.7933532168726033,
"precision": 0.7940298507462686,
"recall": 0.7926777352064709,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9101112629318758,
"precision": 0.9042862684251358,
"recall": 0.9160117878192534,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8923677591320345,
"precision": 0.8961019377214808,
"recall": 0.8886645730521908,
"support": 13374.0
},
"eval_accuracy": 0.848332617245596,
"eval_loss": 0.48704993724823,
"eval_macro avg": {
"f1-score": 0.7983079619318041,
"precision": 0.7982471243223681,
"recall": 0.7983968611807098,
"support": 30257.0
},
"eval_runtime": 4.8769,
"eval_samples_per_second": 16.609,
"eval_steps_per_second": 2.256,
"eval_weighted avg": {
"f1-score": 0.8482045037639816,
"precision": 0.8481162489521332,
"recall": 0.848332617245596,
"support": 30257.0
},
"step": 405
},
{
"epoch": 6.0,
"eval_Claim": {
"f1-score": 0.6060809300245921,
"precision": 0.5903745644599303,
"recall": 0.6226458429030777,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.7858529221479404,
"precision": 0.7963286713286714,
"recall": 0.7756492124308216,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9135615470059578,
"precision": 0.9398504051527115,
"recall": 0.8887033398821218,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.895557603686636,
"precision": 0.88328121591157,
"recall": 0.908180050844923,
"support": 13374.0
},
"eval_accuracy": 0.8502495290346036,
"eval_loss": 0.5423789024353027,
"eval_macro avg": {
"f1-score": 0.8002632507162816,
"precision": 0.8024587142132208,
"recall": 0.7987946115152361,
"support": 30257.0
},
"eval_runtime": 4.867,
"eval_samples_per_second": 16.643,
"eval_steps_per_second": 2.26,
"eval_weighted avg": {
"f1-score": 0.8514422719925407,
"precision": 0.8534139871324166,
"recall": 0.8502495290346036,
"support": 30257.0
},
"step": 486
},
{
"epoch": 6.17,
"grad_norm": 4.494631767272949,
"learning_rate": 1.7530864197530865e-05,
"loss": 0.35,
"step": 500
},
{
"epoch": 7.0,
"eval_Claim": {
"f1-score": 0.6040515653775324,
"precision": 0.6054453161052146,
"recall": 0.6026642168121268,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.8052777186635455,
"precision": 0.8051063829787234,
"recall": 0.8054491272882077,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9184720354599332,
"precision": 0.9315082331548641,
"recall": 0.9057956777996071,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8908606921029282,
"precision": 0.8810881965774463,
"recall": 0.9008524001794527,
"support": 13374.0
},
"eval_accuracy": 0.8521994910268698,
"eval_loss": 0.5764245986938477,
"eval_macro avg": {
"f1-score": 0.8046655029009848,
"precision": 0.8057870322040621,
"recall": 0.8036903555198486,
"support": 30257.0
},
"eval_runtime": 4.8398,
"eval_samples_per_second": 16.736,
"eval_steps_per_second": 2.273,
"eval_weighted avg": {
"f1-score": 0.8522343290465387,
"precision": 0.8524880574571969,
"recall": 0.8521994910268698,
"support": 30257.0
},
"step": 567
},
{
"epoch": 8.0,
"eval_Claim": {
"f1-score": 0.597346719254213,
"precision": 0.6227261400448543,
"recall": 0.5739549839228296,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.8047001620745543,
"precision": 0.7676845767298028,
"recall": 0.8454661558109834,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9099175618286286,
"precision": 0.9258769700050838,
"recall": 0.8944990176817289,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8942491542873952,
"precision": 0.8797569092750688,
"recall": 0.9092268580828473,
"support": 13374.0
},
"eval_accuracy": 0.8510757841160723,
"eval_loss": 0.5954988598823547,
"eval_macro avg": {
"f1-score": 0.8015533993611977,
"precision": 0.7990111490137024,
"recall": 0.8057867538745973,
"support": 30257.0
},
"eval_runtime": 4.8609,
"eval_samples_per_second": 16.663,
"eval_steps_per_second": 2.263,
"eval_weighted avg": {
"f1-score": 0.8498442431569896,
"precision": 0.8495864475589161,
"recall": 0.8510757841160723,
"support": 30257.0
},
"step": 648
},
{
"epoch": 9.0,
"eval_Claim": {
"f1-score": 0.6049033474776048,
"precision": 0.6213075060532688,
"recall": 0.5893431327514929,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.823503227242377,
"precision": 0.8628731343283582,
"recall": 0.787569178373776,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9189658498938837,
"precision": 0.9027672479150872,
"recall": 0.9357563850687622,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8989367655288193,
"precision": 0.8970292606656243,
"recall": 0.9008524001794527,
"support": 13374.0
},
"eval_accuracy": 0.8589747826949136,
"eval_loss": 0.5689579844474792,
"eval_macro avg": {
"f1-score": 0.8115772975356712,
"precision": 0.8209942872405847,
"recall": 0.803380274093371,
"support": 30257.0
},
"eval_runtime": 4.8593,
"eval_samples_per_second": 16.669,
"eval_steps_per_second": 2.264,
"eval_weighted avg": {
"f1-score": 0.8575077142417292,
"precision": 0.8566315824374819,
"recall": 0.8589747826949136,
"support": 30257.0
},
"step": 729
},
{
"epoch": 10.0,
"eval_Claim": {
"f1-score": 0.60162224797219,
"precision": 0.6071094480823199,
"recall": 0.5962333486449242,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.8131676477000218,
"precision": 0.8333333333333334,
"recall": 0.7939548744146445,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9197247706422018,
"precision": 0.9338801134062373,
"recall": 0.905992141453831,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.9008479864909512,
"precision": 0.884834499170693,
"recall": 0.9174517720951099,
"support": 13374.0
},
"eval_accuracy": 0.8577849753775986,
"eval_loss": 0.7088457942008972,
"eval_macro avg": {
"f1-score": 0.8088406632013413,
"precision": 0.8147893484981459,
"recall": 0.8034080341521274,
"support": 30257.0
},
"eval_runtime": 4.9043,
"eval_samples_per_second": 16.516,
"eval_steps_per_second": 2.243,
"eval_weighted avg": {
"f1-score": 0.8573332851434664,
"precision": 0.8573728619273148,
"recall": 0.8577849753775986,
"support": 30257.0
},
"step": 810
},
{
"epoch": 11.0,
"eval_Claim": {
"f1-score": 0.6253176253176252,
"precision": 0.6289498141263941,
"recall": 0.6217271474506201,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.7947071414672116,
"precision": 0.7318996415770609,
"recall": 0.8693060876968923,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9110249527786062,
"precision": 0.9221171261823304,
"recall": 0.900196463654224,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8996578818752584,
"precision": 0.9047258979206049,
"recall": 0.8946463286974727,
"support": 13374.0
},
"eval_accuracy": 0.8552731599299336,
"eval_loss": 0.7312190532684326,
"eval_macro avg": {
"f1-score": 0.8076769003596753,
"precision": 0.7969231199515976,
"recall": 0.8214690068748023,
"support": 30257.0
},
"eval_runtime": 4.9029,
"eval_samples_per_second": 16.521,
"eval_steps_per_second": 2.244,
"eval_weighted avg": {
"f1-score": 0.8558567785115953,
"precision": 0.857475501606145,
"recall": 0.8552731599299336,
"support": 30257.0
},
"step": 891
},
{
"epoch": 12.0,
"eval_Claim": {
"f1-score": 0.633677135241627,
"precision": 0.6395321637426901,
"recall": 0.6279283417547084,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.8290997764681975,
"precision": 0.7931570762052877,
"recall": 0.8684546615581098,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.916678984971668,
"precision": 0.9196243203163619,
"recall": 0.9137524557956778,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.9029209944129889,
"precision": 0.905603610379842,
"recall": 0.9002542246149244,
"support": 13374.0
},
"eval_accuracy": 0.8631391083055161,
"eval_loss": 0.7092207074165344,
"eval_macro avg": {
"f1-score": 0.8205942227736204,
"precision": 0.8144792926610454,
"recall": 0.8275974209308551,
"support": 30257.0
},
"eval_runtime": 4.8734,
"eval_samples_per_second": 16.621,
"eval_steps_per_second": 2.257,
"eval_weighted avg": {
"f1-score": 0.8630744313070275,
"precision": 0.8633032778855295,
"recall": 0.8631391083055161,
"support": 30257.0
},
"step": 972
},
{
"epoch": 12.35,
"grad_norm": 4.426708698272705,
"learning_rate": 1.506172839506173e-05,
"loss": 0.063,
"step": 1000
},
{
"epoch": 13.0,
"eval_Claim": {
"f1-score": 0.5936837143530521,
"precision": 0.6096321393998064,
"recall": 0.5785484611851172,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.8065005417118094,
"precision": 0.821270962047661,
"recall": 0.7922520221370796,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9133826964170352,
"precision": 0.9242683294780247,
"recall": 0.9027504911591355,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.901282521069989,
"precision": 0.883730957171601,
"recall": 0.9195453865709585,
"support": 13374.0
},
"eval_accuracy": 0.854942657897346,
"eval_loss": 0.7807008624076843,
"eval_macro avg": {
"f1-score": 0.8037123683879714,
"precision": 0.8097255970242733,
"recall": 0.7982740902630727,
"support": 30257.0
},
"eval_runtime": 4.8517,
"eval_samples_per_second": 16.695,
"eval_steps_per_second": 2.267,
"eval_weighted avg": {
"f1-score": 0.8537315976828396,
"precision": 0.8530777420132861,
"recall": 0.854942657897346,
"support": 30257.0
},
"step": 1053
},
{
"epoch": 14.0,
"eval_Claim": {
"f1-score": 0.6335702820454024,
"precision": 0.632554945054945,
"recall": 0.6345888837850253,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.8183556405353729,
"precision": 0.816793893129771,
"recall": 0.8199233716475096,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9172396976347231,
"precision": 0.9107990314769976,
"recall": 0.9237721021611002,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8981941309255079,
"precision": 0.9039073148568832,
"recall": 0.8925527142216241,
"support": 13374.0
},
"eval_accuracy": 0.8602967908252636,
"eval_loss": 0.7879041433334351,
"eval_macro avg": {
"f1-score": 0.8168399377852517,
"precision": 0.8160137961296492,
"recall": 0.8177092679538147,
"support": 30257.0
},
"eval_runtime": 4.864,
"eval_samples_per_second": 16.653,
"eval_steps_per_second": 2.262,
"eval_weighted avg": {
"f1-score": 0.8603242501425289,
"precision": 0.8604152313535001,
"recall": 0.8602967908252636,
"support": 30257.0
},
"step": 1134
},
{
"epoch": 15.0,
"eval_Claim": {
"f1-score": 0.6212592513139547,
"precision": 0.5828134433487623,
"recall": 0.6651355075792375,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.7988422575976845,
"precision": 0.7765273311897106,
"recall": 0.822477650063857,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9143625765954267,
"precision": 0.9276255938542404,
"recall": 0.9014734774066797,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8928883984627678,
"precision": 0.9090416053304409,
"recall": 0.8772992373261552,
"support": 13374.0
},
"eval_accuracy": 0.8506461314737086,
"eval_loss": 0.8185116052627563,
"eval_macro avg": {
"f1-score": 0.8068381209924584,
"precision": 0.7990019934307886,
"recall": 0.8165964680939823,
"support": 30257.0
},
"eval_runtime": 4.8575,
"eval_samples_per_second": 16.675,
"eval_steps_per_second": 2.265,
"eval_weighted avg": {
"f1-score": 0.8537245501570023,
"precision": 0.8580620487302317,
"recall": 0.8506461314737086,
"support": 30257.0
},
"step": 1215
},
{
"epoch": 16.0,
"eval_Claim": {
"f1-score": 0.620457750298297,
"precision": 0.5878725590955807,
"recall": 0.6568672485071199,
"support": 4354.0
},
"eval_MajorClaim": {
"f1-score": 0.8217720482913852,
"precision": 0.7911741528762806,
"recall": 0.8548318433375904,
"support": 2349.0
},
"eval_O": {
"f1-score": 0.9171081176529942,
"precision": 0.9427445285758739,
"recall": 0.8928290766208251,
"support": 10180.0
},
"eval_Premise": {
"f1-score": 0.8952495580546884,
"precision": 0.900703852266707,
"recall": 0.8898609241812472,
"support": 13374.0
},
"eval_accuracy": 0.8546121558647586,
"eval_loss": 0.8933361768722534,
"eval_macro avg": {
"f1-score": 0.8136468685743412,
"precision": 0.8056237732036106,
"recall": 0.8235972731616956,
"support": 30257.0
},
"eval_runtime": 4.839,
"eval_samples_per_second": 16.739,
"eval_steps_per_second": 2.273,
"eval_weighted avg": {
"f1-score": 0.8573567707758909,
"precision": 0.8613285464033407,
"recall": 0.8546121558647586,
"support": 30257.0
},
"step": 1296
}
],
"logging_steps": 500,
"max_steps": 4050,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"total_flos": 2293474941878400.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}