longformer-simple / checkpoint-2916 /trainer_state.json
Theoreticallyhugo's picture
Training in progress, epoch 36, checkpoint
9574461 verified
raw
history blame
48.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 36.0,
"eval_steps": 500,
"global_step": 2916,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_Claim": {
"f1-score": 0.41092129347162903,
"precision": 0.5640703517587939,
"recall": 0.32317658349328215,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.6523102655237298,
"precision": 0.5929304446978335,
"recall": 0.724907063197026,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.8834432924913603,
"precision": 0.8545380875202593,
"recall": 0.9143724257533059,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8664166767469054,
"precision": 0.8441231929604023,
"recall": 0.8899196554294707,
"support": 12073.0
},
"eval_accuracy": 0.7997031029363844,
"eval_loss": 0.5104541182518005,
"eval_macro avg": {
"f1-score": 0.7032728820584061,
"precision": 0.7139155192343223,
"recall": 0.7130939319682712,
"support": 27619.0
},
"eval_runtime": 4.7759,
"eval_samples_per_second": 16.751,
"eval_steps_per_second": 2.094,
"eval_weighted avg": {
"f1-score": 0.7866826459135919,
"precision": 0.7857670171690952,
"recall": 0.7997031029363844,
"support": 27619.0
},
"step": 81
},
{
"epoch": 2.0,
"eval_Claim": {
"f1-score": 0.5200312174817897,
"precision": 0.5678977272727272,
"recall": 0.47960652591170827,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.7526302911671151,
"precision": 0.7948320413436692,
"recall": 0.7146840148698885,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9116096466629278,
"precision": 0.944560669456067,
"recall": 0.8808801213960546,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8738735224125151,
"precision": 0.8259587020648967,
"recall": 0.9276898865236478,
"support": 12073.0
},
"eval_accuracy": 0.8278359100619139,
"eval_loss": 0.44692954421043396,
"eval_macro avg": {
"f1-score": 0.7645361694310869,
"precision": 0.7833122850343401,
"recall": 0.7507151371753248,
"support": 27619.0
},
"eval_runtime": 4.7439,
"eval_samples_per_second": 16.864,
"eval_steps_per_second": 2.108,
"eval_weighted avg": {
"f1-score": 0.8236335905447046,
"precision": 0.8242076985653165,
"recall": 0.8278359100619139,
"support": 27619.0
},
"step": 162
},
{
"epoch": 3.0,
"eval_Claim": {
"f1-score": 0.6218357913273168,
"precision": 0.5744204961366409,
"recall": 0.6777831094049904,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.7799868909766223,
"precision": 0.7360824742268042,
"recall": 0.8294609665427509,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9093637121422617,
"precision": 0.9330596419204014,
"recall": 0.8868415347929763,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.876420695504665,
"precision": 0.89797514556357,
"recall": 0.855876749772219,
"support": 12073.0
},
"eval_accuracy": 0.8372859263550454,
"eval_loss": 0.40934404730796814,
"eval_macro avg": {
"f1-score": 0.7969017724877165,
"precision": 0.7853844394618541,
"recall": 0.8124905901282342,
"support": 27619.0
},
"eval_runtime": 4.7941,
"eval_samples_per_second": 16.687,
"eval_steps_per_second": 2.086,
"eval_weighted avg": {
"f1-score": 0.8414917278933443,
"precision": 0.8482528803063183,
"recall": 0.8372859263550454,
"support": 27619.0
},
"step": 243
},
{
"epoch": 4.0,
"eval_Claim": {
"f1-score": 0.608644284729743,
"precision": 0.516112873601603,
"recall": 0.7416026871401151,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.788337924701561,
"precision": 0.7790381125226861,
"recall": 0.7978624535315985,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9068679925629612,
"precision": 0.9442684500762643,
"recall": 0.8723173639713853,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8528029247910863,
"precision": 0.8985600293497202,
"recall": 0.8114801623457302,
"support": 12073.0
},
"eval_accuracy": 0.8201962417176581,
"eval_loss": 0.5160673260688782,
"eval_macro avg": {
"f1-score": 0.789163281696338,
"precision": 0.7844948663875684,
"recall": 0.8058156667472073,
"support": 27619.0
},
"eval_runtime": 4.7669,
"eval_samples_per_second": 16.782,
"eval_steps_per_second": 2.098,
"eval_weighted avg": {
"f1-score": 0.8289940404467937,
"precision": 0.8468005514342695,
"recall": 0.8201962417176581,
"support": 27619.0
},
"step": 324
},
{
"epoch": 5.0,
"eval_Claim": {
"f1-score": 0.5923049770561243,
"precision": 0.5811590856615101,
"recall": 0.6038867562380038,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.7829807463750891,
"precision": 0.8014598540145985,
"recall": 0.7653345724907064,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9075350077795066,
"precision": 0.9311288483466362,
"recall": 0.8851073054411446,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8770785784153896,
"precision": 0.8633555323758325,
"recall": 0.891244926695933,
"support": 12073.0
},
"eval_accuracy": 0.8360186827908324,
"eval_loss": 0.5153732299804688,
"eval_macro avg": {
"f1-score": 0.7899748274065275,
"precision": 0.7942758300996444,
"recall": 0.7863933902164468,
"support": 27619.0
},
"eval_runtime": 4.7506,
"eval_samples_per_second": 16.84,
"eval_steps_per_second": 2.105,
"eval_weighted avg": {
"f1-score": 0.8369451960444566,
"precision": 0.8385857117236715,
"recall": 0.8360186827908324,
"support": 27619.0
},
"step": 405
},
{
"epoch": 6.0,
"eval_Claim": {
"f1-score": 0.5533534584591154,
"precision": 0.6108374384236454,
"recall": 0.5057581573896354,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.7910518800571156,
"precision": 0.8107317073170732,
"recall": 0.7723048327137546,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.909817664129661,
"precision": 0.919331636605068,
"recall": 0.9004985909386516,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8747714081259442,
"precision": 0.8410671966975002,
"recall": 0.9112896546011762,
"support": 12073.0
},
"eval_accuracy": 0.8356566132010572,
"eval_loss": 0.5417770147323608,
"eval_macro avg": {
"f1-score": 0.782248602692959,
"precision": 0.7954919947608217,
"recall": 0.7724628089108043,
"support": 27619.0
},
"eval_runtime": 4.7432,
"eval_samples_per_second": 16.866,
"eval_steps_per_second": 2.108,
"eval_weighted avg": {
"f1-score": 0.8314498656832355,
"precision": 0.8301032992701898,
"recall": 0.8356566132010572,
"support": 27619.0
},
"step": 486
},
{
"epoch": 6.17,
"grad_norm": 6.106111526489258,
"learning_rate": 1.7530864197530865e-05,
"loss": 0.345,
"step": 500
},
{
"epoch": 7.0,
"eval_Claim": {
"f1-score": 0.5860923076923077,
"precision": 0.6017184735911044,
"recall": 0.571257197696737,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8099526066350711,
"precision": 0.8264023210831721,
"recall": 0.7941449814126395,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9184632841627491,
"precision": 0.9157418381639909,
"recall": 0.9212009538261435,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8751742803247764,
"precision": 0.8666450093397222,
"recall": 0.8838731052762362,
"support": 12073.0
},
"eval_accuracy": 0.84217386581701,
"eval_loss": 0.5831862688064575,
"eval_macro avg": {
"f1-score": 0.7974206197037261,
"precision": 0.8026269105444974,
"recall": 0.792619059552939,
"support": 27619.0
},
"eval_runtime": 4.7225,
"eval_samples_per_second": 16.94,
"eval_steps_per_second": 2.118,
"eval_weighted avg": {
"f1-score": 0.8409273360363069,
"precision": 0.8399297508801243,
"recall": 0.84217386581701,
"support": 27619.0
},
"step": 567
},
{
"epoch": 8.0,
"eval_Claim": {
"f1-score": 0.6089751514402275,
"precision": 0.6281560826319816,
"recall": 0.5909309021113244,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8056783802653014,
"precision": 0.806993006993007,
"recall": 0.804368029739777,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9201371876531113,
"precision": 0.9243136825987094,
"recall": 0.9159982657706481,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8842053669893396,
"precision": 0.8721998388396455,
"recall": 0.8965460117617825,
"support": 12073.0
},
"eval_accuracy": 0.8497411202433107,
"eval_loss": 0.5915889739990234,
"eval_macro avg": {
"f1-score": 0.8047490215869949,
"precision": 0.8079156527658359,
"recall": 0.801960802345883,
"support": 27619.0
},
"eval_runtime": 4.7226,
"eval_samples_per_second": 16.94,
"eval_steps_per_second": 2.117,
"eval_weighted avg": {
"f1-score": 0.8485544514458778,
"precision": 0.8476986926907486,
"recall": 0.8497411202433107,
"support": 27619.0
},
"step": 648
},
{
"epoch": 9.0,
"eval_Claim": {
"f1-score": 0.6137689614935823,
"precision": 0.5974557019536574,
"recall": 0.6309980806142035,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8104347826086956,
"precision": 0.761437908496732,
"recall": 0.8661710037174721,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9178336980306346,
"precision": 0.9265518003092555,
"recall": 0.9092781270323,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8752312090129477,
"precision": 0.8886043533930857,
"recall": 0.8622546177420691,
"support": 12073.0
},
"eval_accuracy": 0.8433686954632681,
"eval_loss": 0.6680669784545898,
"eval_macro avg": {
"f1-score": 0.804317162786465,
"precision": 0.7935124410381827,
"recall": 0.8171754572765112,
"support": 27619.0
},
"eval_runtime": 4.7195,
"eval_samples_per_second": 16.951,
"eval_steps_per_second": 2.119,
"eval_weighted avg": {
"f1-score": 0.8449561811840803,
"precision": 0.8474346288061019,
"recall": 0.8433686954632681,
"support": 27619.0
},
"step": 729
},
{
"epoch": 10.0,
"eval_Claim": {
"f1-score": 0.5938368860055606,
"precision": 0.5741487455197133,
"recall": 0.6149232245681382,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8202093764223942,
"precision": 0.8037466547725245,
"recall": 0.837360594795539,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9145026408450704,
"precision": 0.9286033519553073,
"recall": 0.9008237589421201,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8716092527874855,
"precision": 0.8756164841594918,
"recall": 0.8676385322620724,
"support": 12073.0
},
"eval_accuracy": 0.8382273072884608,
"eval_loss": 0.7243372201919556,
"eval_macro avg": {
"f1-score": 0.8000395390151277,
"precision": 0.7955288091017592,
"recall": 0.8051865276419675,
"support": 27619.0
},
"eval_runtime": 4.7291,
"eval_samples_per_second": 16.917,
"eval_steps_per_second": 2.115,
"eval_weighted avg": {
"f1-score": 0.840013852502701,
"precision": 0.8422219164630815,
"recall": 0.8382273072884608,
"support": 27619.0
},
"step": 810
},
{
"epoch": 11.0,
"eval_Claim": {
"f1-score": 0.6018801410105759,
"precision": 0.5898203592814372,
"recall": 0.614443378119002,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.7984663311766117,
"precision": 0.8243443839683325,
"recall": 0.7741635687732342,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9056203605514316,
"precision": 0.9334944195144402,
"recall": 0.8793626707132018,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.878886273236464,
"precision": 0.8616792678074016,
"recall": 0.8967945001242442,
"support": 12073.0
},
"eval_accuracy": 0.838806618632101,
"eval_loss": 0.7492838501930237,
"eval_macro avg": {
"f1-score": 0.7962132764937708,
"precision": 0.8023346076429029,
"recall": 0.7911910294324205,
"support": 27619.0
},
"eval_runtime": 4.7534,
"eval_samples_per_second": 16.83,
"eval_steps_per_second": 2.104,
"eval_weighted avg": {
"f1-score": 0.8397473983726959,
"precision": 0.8417333606018996,
"recall": 0.838806618632101,
"support": 27619.0
},
"step": 891
},
{
"epoch": 12.0,
"eval_Claim": {
"f1-score": 0.6351054481546573,
"precision": 0.5856969205834684,
"recall": 0.6936180422264875,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8096330275229358,
"precision": 0.7993659420289855,
"recall": 0.8201672862453532,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9158836689038031,
"precision": 0.9461520684076727,
"recall": 0.8874918707999133,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8832342847576798,
"precision": 0.8926486760849336,
"recall": 0.8740164002319225,
"support": 12073.0
},
"eval_accuracy": 0.8470980122379521,
"eval_loss": 0.7472469806671143,
"eval_macro avg": {
"f1-score": 0.810964107334769,
"precision": 0.8059659017762651,
"recall": 0.8188233998759191,
"support": 27619.0
},
"eval_runtime": 4.7526,
"eval_samples_per_second": 16.833,
"eval_steps_per_second": 2.104,
"eval_weighted avg": {
"f1-score": 0.8509605717920246,
"precision": 0.8569306173916821,
"recall": 0.8470980122379521,
"support": 27619.0
},
"step": 972
},
{
"epoch": 12.35,
"grad_norm": 3.1761698722839355,
"learning_rate": 1.506172839506173e-05,
"loss": 0.0604,
"step": 1000
},
{
"epoch": 13.0,
"eval_Claim": {
"f1-score": 0.6302578018995931,
"precision": 0.5960222412318221,
"recall": 0.6686660268714012,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8086405259450574,
"precision": 0.8172757475083057,
"recall": 0.800185873605948,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9224348987313599,
"precision": 0.9477475417333638,
"recall": 0.8984391935833514,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8844928196002152,
"precision": 0.8838709677419355,
"recall": 0.8851155470885447,
"support": 12073.0
},
"eval_accuracy": 0.8502842246279735,
"eval_loss": 0.812764048576355,
"eval_macro avg": {
"f1-score": 0.8114565115440564,
"precision": 0.8112291245538568,
"recall": 0.8131016602873112,
"support": 27619.0
},
"eval_runtime": 4.7412,
"eval_samples_per_second": 16.873,
"eval_steps_per_second": 2.109,
"eval_weighted avg": {
"f1-score": 0.8528902247322565,
"precision": 0.8565802934093382,
"recall": 0.8502842246279735,
"support": 27619.0
},
"step": 1053
},
{
"epoch": 14.0,
"eval_Claim": {
"f1-score": 0.6177896267571498,
"precision": 0.6241429970617042,
"recall": 0.6115642994241842,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.7891914191419142,
"precision": 0.7095697329376854,
"recall": 0.8889405204460966,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.915085817524842,
"precision": 0.9549846806504831,
"recall": 0.8783871667027965,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8833210513387375,
"precision": 0.8733101270946328,
"recall": 0.8935641514122422,
"support": 12073.0
},
"eval_accuracy": 0.8455773199608965,
"eval_loss": 0.8994765281677246,
"eval_macro avg": {
"f1-score": 0.8013469786906608,
"precision": 0.7905018844361265,
"recall": 0.8181140344963299,
"support": 27619.0
},
"eval_runtime": 4.7538,
"eval_samples_per_second": 16.829,
"eval_steps_per_second": 2.104,
"eval_weighted avg": {
"f1-score": 0.8465260836240982,
"precision": 0.850232952139105,
"recall": 0.8455773199608965,
"support": 27619.0
},
"step": 1134
},
{
"epoch": 15.0,
"eval_Claim": {
"f1-score": 0.6071797224452345,
"precision": 0.5816304108987036,
"recall": 0.6350767754318618,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8077178975382568,
"precision": 0.7725922783198982,
"recall": 0.8461895910780669,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9198298062926884,
"precision": 0.9512505789717461,
"recall": 0.8904183828311294,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8720390922643698,
"precision": 0.8719668737060041,
"recall": 0.8721113227863828,
"support": 12073.0
},
"eval_accuracy": 0.8404359317860893,
"eval_loss": 0.9128327369689941,
"eval_macro avg": {
"f1-score": 0.8016916296351374,
"precision": 0.794360035474088,
"recall": 0.8109490180318603,
"support": 27619.0
},
"eval_runtime": 4.7243,
"eval_samples_per_second": 16.934,
"eval_steps_per_second": 2.117,
"eval_weighted avg": {
"f1-score": 0.8430215341764057,
"precision": 0.8468933720777774,
"recall": 0.8404359317860893,
"support": 27619.0
},
"step": 1215
},
{
"epoch": 16.0,
"eval_Claim": {
"f1-score": 0.6309481002037637,
"precision": 0.6304191616766467,
"recall": 0.6314779270633397,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8244803695150116,
"precision": 0.8195592286501377,
"recall": 0.8294609665427509,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9188405797101449,
"precision": 0.9101446193109315,
"recall": 0.9277043138955127,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8854815124294965,
"precision": 0.8933569381217332,
"recall": 0.8777437256688478,
"support": 12073.0
},
"eval_accuracy": 0.8535066439769724,
"eval_loss": 0.8802785873413086,
"eval_macro avg": {
"f1-score": 0.8149376404646043,
"precision": 0.8133699869398623,
"recall": 0.8165967332926127,
"support": 27619.0
},
"eval_runtime": 4.7225,
"eval_samples_per_second": 16.94,
"eval_steps_per_second": 2.118,
"eval_weighted avg": {
"f1-score": 0.8534601153123827,
"precision": 0.8535345630772185,
"recall": 0.8535066439769724,
"support": 27619.0
},
"step": 1296
},
{
"epoch": 17.0,
"eval_Claim": {
"f1-score": 0.6351891003053795,
"precision": 0.6221813161527842,
"recall": 0.6487523992322457,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8157058556513844,
"precision": 0.7972493345164152,
"recall": 0.8350371747211895,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9181362943609842,
"precision": 0.9221517603323858,
"recall": 0.914155647084327,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8847406664996242,
"precision": 0.8921923692411353,
"recall": 0.8774124078522323,
"support": 12073.0
},
"eval_accuracy": 0.8518773308229842,
"eval_loss": 0.8617361783981323,
"eval_macro avg": {
"f1-score": 0.813442979204343,
"precision": 0.8084436950606801,
"recall": 0.8188394072224987,
"support": 27619.0
},
"eval_runtime": 4.7584,
"eval_samples_per_second": 16.812,
"eval_steps_per_second": 2.102,
"eval_weighted avg": {
"f1-score": 0.8528573333523664,
"precision": 0.8540549226358285,
"recall": 0.8518773308229842,
"support": 27619.0
},
"step": 1377
},
{
"epoch": 18.0,
"eval_Claim": {
"f1-score": 0.5979982262764474,
"precision": 0.6335570469798658,
"recall": 0.5662188099808061,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8218923933209649,
"precision": 0.8203703703703704,
"recall": 0.8234200743494424,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9185111222055806,
"precision": 0.9406885581183957,
"recall": 0.8973553002384566,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8841078141246101,
"precision": 0.8547127503286167,
"recall": 0.9155967862171789,
"support": 12073.0
},
"eval_accuracy": 0.8495962924074008,
"eval_loss": 0.9965270757675171,
"eval_macro avg": {
"f1-score": 0.8056273889819007,
"precision": 0.8123321814493122,
"recall": 0.800647742696471,
"support": 27619.0
},
"eval_runtime": 4.735,
"eval_samples_per_second": 16.895,
"eval_steps_per_second": 2.112,
"eval_weighted avg": {
"f1-score": 0.8475754477331567,
"precision": 0.8473819646173586,
"recall": 0.8495962924074008,
"support": 27619.0
},
"step": 1458
},
{
"epoch": 18.52,
"grad_norm": 15.414011001586914,
"learning_rate": 1.2592592592592593e-05,
"loss": 0.0201,
"step": 1500
},
{
"epoch": 19.0,
"eval_Claim": {
"f1-score": 0.6417134539405281,
"precision": 0.5662385321100918,
"recall": 0.7404030710172744,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8129032258064516,
"precision": 0.8062157221206582,
"recall": 0.8197026022304833,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9158699808795411,
"precision": 0.9517297802711547,
"recall": 0.8826143507478864,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8692654694016513,
"precision": 0.8939168490153173,
"recall": 0.8459372152737513,
"support": 12073.0
},
"eval_accuracy": 0.8402186900322242,
"eval_loss": 0.9760411381721497,
"eval_macro avg": {
"f1-score": 0.809938032507043,
"precision": 0.8045252208793054,
"recall": 0.8221643098173489,
"support": 27619.0
},
"eval_runtime": 4.7957,
"eval_samples_per_second": 16.682,
"eval_steps_per_second": 2.085,
"eval_weighted avg": {
"f1-score": 0.8461018818074655,
"precision": 0.8569454182549012,
"recall": 0.8402186900322242,
"support": 27619.0
},
"step": 1539
},
{
"epoch": 20.0,
"eval_Claim": {
"f1-score": 0.651902203705734,
"precision": 0.6002422774076317,
"recall": 0.7132917466410749,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8352472463088821,
"precision": 0.8425531914893617,
"recall": 0.8280669144981413,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9225629791894853,
"precision": 0.9323666150099623,
"recall": 0.9129633644049425,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8775752437473505,
"precision": 0.8987583572110793,
"recall": 0.8573676799469891,
"support": 12073.0
},
"eval_accuracy": 0.8519135377819617,
"eval_loss": 0.9999569058418274,
"eval_macro avg": {
"f1-score": 0.821821918237863,
"precision": 0.8184801102795087,
"recall": 0.8279224263727869,
"support": 27619.0
},
"eval_runtime": 4.7617,
"eval_samples_per_second": 16.801,
"eval_steps_per_second": 2.1,
"eval_weighted avg": {
"f1-score": 0.8552486484979966,
"precision": 0.8605564400235849,
"recall": 0.8519135377819617,
"support": 27619.0
},
"step": 1620
},
{
"epoch": 21.0,
"eval_Claim": {
"f1-score": 0.6376654371881102,
"precision": 0.5819801980198019,
"recall": 0.7051343570057581,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.792661361626879,
"precision": 0.7559021922428331,
"recall": 0.83317843866171,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9065665187208377,
"precision": 0.9542405366554864,
"recall": 0.8634294385432474,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8798595435164284,
"precision": 0.8881762174023125,
"recall": 0.8716971755156133,
"support": 12073.0
},
"eval_accuracy": 0.8407980013758645,
"eval_loss": 1.0823462009429932,
"eval_macro avg": {
"f1-score": 0.8041882152630638,
"precision": 0.7950747860801084,
"recall": 0.8183598524315823,
"support": 27619.0
},
"eval_runtime": 4.8337,
"eval_samples_per_second": 16.55,
"eval_steps_per_second": 2.069,
"eval_weighted avg": {
"f1-score": 0.8454369732073347,
"precision": 0.8537300281304444,
"recall": 0.8407980013758645,
"support": 27619.0
},
"step": 1701
},
{
"epoch": 22.0,
"eval_Claim": {
"f1-score": 0.6413618587531631,
"precision": 0.6159964648696421,
"recall": 0.6689059500959693,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8162887552059233,
"precision": 0.8129032258064516,
"recall": 0.8197026022304833,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9215077605321508,
"precision": 0.9430451554345359,
"recall": 0.9009321482766096,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8851211645025227,
"precision": 0.8838054339747295,
"recall": 0.886440818355007,
"support": 12073.0
},
"eval_accuracy": 0.8532531952641298,
"eval_loss": 0.9629964828491211,
"eval_macro avg": {
"f1-score": 0.8160698847484399,
"precision": 0.8139375700213398,
"recall": 0.8189953797395173,
"support": 27619.0
},
"eval_runtime": 4.7476,
"eval_samples_per_second": 16.851,
"eval_steps_per_second": 2.106,
"eval_weighted avg": {
"f1-score": 0.8551268346498754,
"precision": 0.8576544630844016,
"recall": 0.8532531952641298,
"support": 27619.0
},
"step": 1782
},
{
"epoch": 23.0,
"eval_Claim": {
"f1-score": 0.6246913580246913,
"precision": 0.6434384537131231,
"recall": 0.6070057581573897,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8306414397784957,
"precision": 0.8249312557286893,
"recall": 0.8364312267657993,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9210203854294309,
"precision": 0.9472960586617782,
"recall": 0.8961630175590722,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8880482897384305,
"precision": 0.8635830007043908,
"recall": 0.9139401971341009,
"support": 12073.0
},
"eval_accuracy": 0.8556428545566458,
"eval_loss": 1.0519344806671143,
"eval_macro avg": {
"f1-score": 0.8161003682427621,
"precision": 0.8198121922019954,
"recall": 0.8133850499040906,
"support": 27619.0
},
"eval_runtime": 4.7514,
"eval_samples_per_second": 16.837,
"eval_steps_per_second": 2.105,
"eval_weighted avg": {
"f1-score": 0.8548461217507599,
"precision": 0.8553131736167897,
"recall": 0.8556428545566458,
"support": 27619.0
},
"step": 1863
},
{
"epoch": 24.0,
"eval_Claim": {
"f1-score": 0.6429456142326315,
"precision": 0.605771270952684,
"recall": 0.6849808061420346,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8262760357061112,
"precision": 0.8141632837167343,
"recall": 0.8387546468401487,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9164846244328684,
"precision": 0.9483018430508867,
"recall": 0.8867331454584869,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8827014708928941,
"precision": 0.8831039628585641,
"recall": 0.8822993456473122,
"support": 12073.0
},
"eval_accuracy": 0.8506100872587712,
"eval_loss": 1.0345954895019531,
"eval_macro avg": {
"f1-score": 0.8171019363161263,
"precision": 0.8128350901447173,
"recall": 0.8231919860219956,
"support": 27619.0
},
"eval_runtime": 4.7433,
"eval_samples_per_second": 16.866,
"eval_steps_per_second": 2.108,
"eval_weighted avg": {
"f1-score": 0.8534083548306859,
"precision": 0.8576588939305595,
"recall": 0.8506100872587712,
"support": 27619.0
},
"step": 1944
},
{
"epoch": 24.69,
"grad_norm": 0.018422244116663933,
"learning_rate": 1.0123456790123458e-05,
"loss": 0.0091,
"step": 2000
},
{
"epoch": 25.0,
"eval_Claim": {
"f1-score": 0.6397129186602871,
"precision": 0.637881679389313,
"recall": 0.6415547024952015,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8255892255892255,
"precision": 0.798523664785063,
"recall": 0.8545539033457249,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9217740138176956,
"precision": 0.9484063288236643,
"recall": 0.8965965748970302,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8884167517875383,
"precision": 0.8766328011611031,
"recall": 0.9005218255611696,
"support": 12073.0
},
"eval_accuracy": 0.8565480285310837,
"eval_loss": 1.0353806018829346,
"eval_macro avg": {
"f1-score": 0.8188732274636866,
"precision": 0.8153611185397858,
"recall": 0.8233067515747815,
"support": 27619.0
},
"eval_runtime": 4.7632,
"eval_samples_per_second": 16.795,
"eval_steps_per_second": 2.099,
"eval_weighted avg": {
"f1-score": 0.8571321899510156,
"precision": 0.8584922830101466,
"recall": 0.8565480285310837,
"support": 27619.0
},
"step": 2025
},
{
"epoch": 26.0,
"eval_Claim": {
"f1-score": 0.639289678135405,
"precision": 0.6577010910936311,
"recall": 0.6218809980806143,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8237179487179487,
"precision": 0.8118231046931408,
"recall": 0.8359665427509294,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9226105170323007,
"precision": 0.9386496186630776,
"recall": 0.9071103403425103,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8938624639506073,
"precision": 0.8770125936553483,
"recall": 0.9113724840553301,
"support": 12073.0
},
"eval_accuracy": 0.8603859661827004,
"eval_loss": 1.0636707544326782,
"eval_macro avg": {
"f1-score": 0.8198701519590654,
"precision": 0.8212966020262995,
"recall": 0.8190825913073461,
"support": 27619.0
},
"eval_runtime": 4.7618,
"eval_samples_per_second": 16.8,
"eval_steps_per_second": 2.1,
"eval_weighted avg": {
"f1-score": 0.8595824092662689,
"precision": 0.8594263331027361,
"recall": 0.8603859661827004,
"support": 27619.0
},
"step": 2106
},
{
"epoch": 27.0,
"eval_Claim": {
"f1-score": 0.643669985775249,
"precision": 0.6361293345829429,
"recall": 0.6513915547024952,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.835773822933093,
"precision": 0.8111062527328378,
"recall": 0.8619888475836431,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9197544642857143,
"precision": 0.9478951000690131,
"recall": 0.893236505527856,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.892443644397169,
"precision": 0.8817299919159256,
"recall": 0.9034208564565559,
"support": 12073.0
},
"eval_accuracy": 0.8587566530287121,
"eval_loss": 1.0722382068634033,
"eval_macro avg": {
"f1-score": 0.8229104793478063,
"precision": 0.8192151698251797,
"recall": 0.8275094410676376,
"support": 27619.0
},
"eval_runtime": 4.7416,
"eval_samples_per_second": 16.872,
"eval_steps_per_second": 2.109,
"eval_weighted avg": {
"f1-score": 0.8596085511412532,
"precision": 0.8612655819566406,
"recall": 0.8587566530287121,
"support": 27619.0
},
"step": 2187
},
{
"epoch": 28.0,
"eval_Claim": {
"f1-score": 0.6362290227048372,
"precision": 0.654979674796748,
"recall": 0.6185220729366603,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.831056793673616,
"precision": 0.8579910935180604,
"recall": 0.8057620817843866,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9264762322495751,
"precision": 0.9374237212914679,
"recall": 0.9157814871016692,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8929698244478602,
"precision": 0.8726381532136928,
"recall": 0.9142715149507165,
"support": 12073.0
},
"eval_accuracy": 0.8616894167058908,
"eval_loss": 1.0252758264541626,
"eval_macro avg": {
"f1-score": 0.8216829682689721,
"precision": 0.8307581607049922,
"recall": 0.8135842891933581,
"support": 27619.0
},
"eval_runtime": 4.7454,
"eval_samples_per_second": 16.858,
"eval_steps_per_second": 2.107,
"eval_weighted avg": {
"f1-score": 0.8605934753580137,
"precision": 0.8602912413261778,
"recall": 0.8616894167058908,
"support": 27619.0
},
"step": 2268
},
{
"epoch": 29.0,
"eval_Claim": {
"f1-score": 0.6346723545443773,
"precision": 0.6270194333879654,
"recall": 0.6425143953934741,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8310436249698723,
"precision": 0.8632949424136205,
"recall": 0.8011152416356877,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9246601728028178,
"precision": 0.9391839016210173,
"recall": 0.9105787990461739,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8919482004983863,
"precision": 0.8799774302756731,
"recall": 0.9042491509980949,
"support": 12073.0
},
"eval_accuracy": 0.8588290669466672,
"eval_loss": 1.0103356838226318,
"eval_macro avg": {
"f1-score": 0.8205810882038633,
"precision": 0.8273689269245691,
"recall": 0.8146143967683577,
"support": 27619.0
},
"eval_runtime": 4.7642,
"eval_samples_per_second": 16.792,
"eval_steps_per_second": 2.099,
"eval_weighted avg": {
"f1-score": 0.8593043062229604,
"precision": 0.8602811798583894,
"recall": 0.8588290669466672,
"support": 27619.0
},
"step": 2349
},
{
"epoch": 30.0,
"eval_Claim": {
"f1-score": 0.6266573621772505,
"precision": 0.6081264108352145,
"recall": 0.6463531669865643,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8181200453001133,
"precision": 0.7980556783031374,
"recall": 0.8392193308550185,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9268776883202824,
"precision": 0.9434216434665469,
"recall": 0.9109039670496423,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8838985513262215,
"precision": 0.8859211183225162,
"recall": 0.8818851983765427,
"support": 12073.0
},
"eval_accuracy": 0.852710090879467,
"eval_loss": 1.0344375371932983,
"eval_macro avg": {
"f1-score": 0.8138884117809669,
"precision": 0.8088812127318536,
"recall": 0.819590415816942,
"support": 27619.0
},
"eval_runtime": 4.7499,
"eval_samples_per_second": 16.843,
"eval_steps_per_second": 2.105,
"eval_weighted avg": {
"f1-score": 0.8543098224282204,
"precision": 0.8563604925666979,
"recall": 0.852710090879467,
"support": 27619.0
},
"step": 2430
},
{
"epoch": 30.86,
"grad_norm": 0.13328897953033447,
"learning_rate": 7.654320987654322e-06,
"loss": 0.0054,
"step": 2500
},
{
"epoch": 31.0,
"eval_Claim": {
"f1-score": 0.6430961409770243,
"precision": 0.6008753647353063,
"recall": 0.6916986564299424,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8156321839080459,
"precision": 0.8070973612374887,
"recall": 0.8243494423791822,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.922759926491062,
"precision": 0.948917649753751,
"recall": 0.8980056362453934,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.884039223868141,
"precision": 0.8907669021190716,
"recall": 0.8774124078522323,
"support": 12073.0
},
"eval_accuracy": 0.8521307795358268,
"eval_loss": 1.0802738666534424,
"eval_macro avg": {
"f1-score": 0.8163818688110683,
"precision": 0.8119143194614044,
"recall": 0.8228665357266876,
"support": 27619.0
},
"eval_runtime": 4.7458,
"eval_samples_per_second": 16.857,
"eval_steps_per_second": 2.107,
"eval_weighted avg": {
"f1-score": 0.85528273315214,
"precision": 0.8599248737286467,
"recall": 0.8521307795358268,
"support": 27619.0
},
"step": 2511
},
{
"epoch": 32.0,
"eval_Claim": {
"f1-score": 0.6231684408265498,
"precision": 0.6518208016767094,
"recall": 0.5969289827255279,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.802954491303312,
"precision": 0.823960880195599,
"recall": 0.7829925650557621,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9214892082339233,
"precision": 0.9439581675571217,
"recall": 0.9000650336006937,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8923421084168898,
"precision": 0.8618055555555556,
"recall": 0.925122173444877,
"support": 12073.0
},
"eval_accuracy": 0.856149751982331,
"eval_loss": 1.0980205535888672,
"eval_macro avg": {
"f1-score": 0.8099885621951687,
"precision": 0.8203863512462465,
"recall": 0.8012771887067152,
"support": 27619.0
},
"eval_runtime": 4.7314,
"eval_samples_per_second": 16.908,
"eval_steps_per_second": 2.114,
"eval_weighted avg": {
"f1-score": 0.8544925535585312,
"precision": 0.8546105739408988,
"recall": 0.856149751982331,
"support": 27619.0
},
"step": 2592
},
{
"epoch": 33.0,
"eval_Claim": {
"f1-score": 0.645293315143247,
"precision": 0.6132238547968885,
"recall": 0.6809021113243762,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8171494785631519,
"precision": 0.8150716597318539,
"recall": 0.8192379182156134,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9240913120567376,
"precision": 0.9452505100884153,
"recall": 0.9038586603078257,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8876614477345405,
"precision": 0.8901382642012328,
"recall": 0.8851983765426986,
"support": 12073.0
},
"eval_accuracy": 0.8554618197617582,
"eval_loss": 1.103171467781067,
"eval_macro avg": {
"f1-score": 0.8185488883744192,
"precision": 0.8159210722045976,
"recall": 0.8222992665976285,
"support": 27619.0
},
"eval_runtime": 4.7195,
"eval_samples_per_second": 16.951,
"eval_steps_per_second": 2.119,
"eval_weighted avg": {
"f1-score": 0.8577606473413056,
"precision": 0.8609099427319448,
"recall": 0.8554618197617582,
"support": 27619.0
},
"step": 2673
},
{
"epoch": 34.0,
"eval_Claim": {
"f1-score": 0.6230995586071605,
"precision": 0.6371614844533601,
"recall": 0.6096449136276392,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8227470763586332,
"precision": 0.812132186509733,
"recall": 0.8336431226765799,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9239675800849093,
"precision": 0.9402985074626866,
"recall": 0.9081942336874052,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8910673608851285,
"precision": 0.875469586763648,
"recall": 0.9072310113476352,
"support": 12073.0
},
"eval_accuracy": 0.8569100981208588,
"eval_loss": 1.1183395385742188,
"eval_macro avg": {
"f1-score": 0.8152203939839578,
"precision": 0.8162654412973569,
"recall": 0.8146783203348149,
"support": 27619.0
},
"eval_runtime": 4.7358,
"eval_samples_per_second": 16.893,
"eval_steps_per_second": 2.112,
"eval_weighted avg": {
"f1-score": 0.8562950074379215,
"precision": 0.856227085825657,
"recall": 0.8569100981208588,
"support": 27619.0
},
"step": 2754
},
{
"epoch": 35.0,
"eval_Claim": {
"f1-score": 0.6287781538102462,
"precision": 0.6121336059986366,
"recall": 0.6463531669865643,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8306414397784957,
"precision": 0.8249312557286893,
"recall": 0.8364312267657993,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9218662169758292,
"precision": 0.9574964969640355,
"recall": 0.8887925428137872,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8870238337746996,
"precision": 0.8728351507376524,
"recall": 0.9016814379193241,
"support": 12073.0
},
"eval_accuracy": 0.853760092689815,
"eval_loss": 1.126227617263794,
"eval_macro avg": {
"f1-score": 0.8170774110848177,
"precision": 0.8168491273572535,
"recall": 0.8183145936213687,
"support": 27619.0
},
"eval_runtime": 4.7324,
"eval_samples_per_second": 16.905,
"eval_steps_per_second": 2.113,
"eval_weighted avg": {
"f1-score": 0.8552975917471803,
"precision": 0.8580407106693336,
"recall": 0.853760092689815,
"support": 27619.0
},
"step": 2835
},
{
"epoch": 36.0,
"eval_Claim": {
"f1-score": 0.6262910798122067,
"precision": 0.6130514705882353,
"recall": 0.6401151631477927,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8252405459834415,
"precision": 0.7958567112645663,
"recall": 0.8568773234200744,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9250767085076709,
"precision": 0.9530980572479595,
"recall": 0.8986559722523304,
"support": 9226.0
},
"eval_Premise": {
"f1-score": 0.8858740338760073,
"precision": 0.87943841319076,
"recall": 0.8924045390540877,
"support": 12073.0
},
"eval_accuracy": 0.8536514718128825,
"eval_loss": 1.135550856590271,
"eval_macro avg": {
"f1-score": 0.8156205920448316,
"precision": 0.8103611630728802,
"recall": 0.8220132494685712,
"support": 27619.0
},
"eval_runtime": 4.7408,
"eval_samples_per_second": 16.875,
"eval_steps_per_second": 2.109,
"eval_weighted avg": {
"f1-score": 0.8550712842351443,
"precision": 0.8573309971640839,
"recall": 0.8536514718128825,
"support": 27619.0
},
"step": 2916
}
],
"logging_steps": 500,
"max_steps": 4050,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"total_flos": 5176394378164800.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}