longformer-spans / checkpoint-1620 /trainer_state.json
Theoreticallyhugo's picture
Training in progress, epoch 20, checkpoint
c2b56bc verified
raw
history blame
23.4 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 20.0,
"eval_steps": 500,
"global_step": 1620,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_B": {
"f1-score": 0.8266764382557715,
"precision": 0.7866108786610879,
"recall": 0.871042471042471,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9410658617447671,
"precision": 0.9344077040239768,
"recall": 0.9478195863443808,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.8562443302086484,
"precision": 0.8789421405512788,
"recall": 0.8346893055064262,
"support": 8481.0
},
"eval_accuracy": 0.912335377500754,
"eval_loss": 0.2450982630252838,
"eval_macro avg": {
"f1-score": 0.8746622100697291,
"precision": 0.8666535744121145,
"recall": 0.8845171209644261,
"support": 29841.0
},
"eval_runtime": 4.6269,
"eval_samples_per_second": 17.29,
"eval_steps_per_second": 2.161,
"eval_weighted avg": {
"f1-score": 0.9119949287205363,
"precision": 0.9122301519092053,
"recall": 0.912335377500754,
"support": 29841.0
},
"step": 81
},
{
"epoch": 2.0,
"eval_B": {
"f1-score": 0.8833395452851286,
"precision": 0.8537463976945245,
"recall": 0.915057915057915,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9549518227873249,
"precision": 0.9516903430183636,
"recall": 0.958235733864939,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.88924750463212,
"precision": 0.9016969696969697,
"recall": 0.877137130055418,
"support": 8481.0
},
"eval_accuracy": 0.9333132267685399,
"eval_loss": 0.20076116919517517,
"eval_macro avg": {
"f1-score": 0.9091796242348579,
"precision": 0.9023779034699526,
"recall": 0.916810259659424,
"support": 29841.0
},
"eval_runtime": 4.6524,
"eval_samples_per_second": 17.195,
"eval_steps_per_second": 2.149,
"eval_weighted avg": {
"f1-score": 0.9331705077630417,
"precision": 0.9332314707173981,
"recall": 0.9333132267685399,
"support": 29841.0
},
"step": 162
},
{
"epoch": 3.0,
"eval_B": {
"f1-score": 0.8927335640138409,
"precision": 0.8889739663093414,
"recall": 0.8965250965250965,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9590235396687009,
"precision": 0.9586653386454184,
"recall": 0.9593820084724645,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9006849315068494,
"precision": 0.9020697811945594,
"recall": 0.8993043273198915,
"support": 8481.0
},
"eval_accuracy": 0.9395797727958178,
"eval_loss": 0.19863653182983398,
"eval_macro avg": {
"f1-score": 0.9174806783964637,
"precision": 0.9165696953831063,
"recall": 0.9184038107724842,
"support": 29841.0
},
"eval_runtime": 4.655,
"eval_samples_per_second": 17.186,
"eval_steps_per_second": 2.148,
"eval_weighted avg": {
"f1-score": 0.9395665759512081,
"precision": 0.9395561516236713,
"recall": 0.9395797727958178,
"support": 29841.0
},
"step": 243
},
{
"epoch": 4.0,
"eval_B": {
"f1-score": 0.8999616711383672,
"precision": 0.893455098934551,
"recall": 0.9065637065637066,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9612657113512985,
"precision": 0.9436362763454799,
"recall": 0.9795664091701969,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.8996847765622102,
"precision": 0.9454403741231489,
"recall": 0.8581535196321188,
"support": 8481.0
},
"eval_accuracy": 0.9418920277470594,
"eval_loss": 0.19743724167346954,
"eval_macro avg": {
"f1-score": 0.920304053017292,
"precision": 0.9275105831343932,
"recall": 0.9147612117886741,
"support": 29841.0
},
"eval_runtime": 4.6615,
"eval_samples_per_second": 17.162,
"eval_steps_per_second": 2.145,
"eval_weighted avg": {
"f1-score": 0.9411036309913238,
"precision": 0.9419713163409645,
"recall": 0.9418920277470594,
"support": 29841.0
},
"step": 324
},
{
"epoch": 5.0,
"eval_B": {
"f1-score": 0.8953846153846153,
"precision": 0.8919540229885058,
"recall": 0.8988416988416988,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9602426713742261,
"precision": 0.9581224570804803,
"recall": 0.9623722900573137,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9022119433078337,
"precision": 0.9075399665950847,
"recall": 0.8969461148449476,
"support": 8481.0
},
"eval_accuracy": 0.9410207432726785,
"eval_loss": 0.2340461015701294,
"eval_macro avg": {
"f1-score": 0.9192797433555584,
"precision": 0.9192054822213569,
"recall": 0.9193867012479867,
"support": 29841.0
},
"eval_runtime": 4.6719,
"eval_samples_per_second": 17.124,
"eval_steps_per_second": 2.14,
"eval_weighted avg": {
"f1-score": 0.9409353496612266,
"precision": 0.9408751053176122,
"recall": 0.9410207432726785,
"support": 29841.0
},
"step": 405
},
{
"epoch": 6.0,
"eval_B": {
"f1-score": 0.8929831438651511,
"precision": 0.9068471337579618,
"recall": 0.8795366795366796,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9522293798155866,
"precision": 0.9655225409836066,
"recall": 0.9392972838275604,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.8885216003647556,
"precision": 0.8599007170435742,
"recall": 0.919113312109421,
"support": 8481.0
},
"eval_accuracy": 0.930967460875976,
"eval_loss": 0.21774999797344208,
"eval_macro avg": {
"f1-score": 0.9112447080151643,
"precision": 0.9107567972617142,
"recall": 0.9126490918245537,
"support": 29841.0
},
"eval_runtime": 4.647,
"eval_samples_per_second": 17.215,
"eval_steps_per_second": 2.152,
"eval_weighted avg": {
"f1-score": 0.9315521386682287,
"precision": 0.9329578366776978,
"recall": 0.930967460875976,
"support": 29841.0
},
"step": 486
},
{
"epoch": 6.17,
"grad_norm": 9.310327529907227,
"learning_rate": 1.7530864197530865e-05,
"loss": 0.1627,
"step": 500
},
{
"epoch": 7.0,
"eval_B": {
"f1-score": 0.8936490041337843,
"precision": 0.8704245973645681,
"recall": 0.9181467181467181,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9599271850630982,
"precision": 0.9477800446905664,
"recall": 0.9723897333665587,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.8979841172877214,
"precision": 0.9316770186335404,
"recall": 0.8666430845419172,
"support": 8481.0
},
"eval_accuracy": 0.9399819040916859,
"eval_loss": 0.27376389503479004,
"eval_macro avg": {
"f1-score": 0.9171867688282013,
"precision": 0.9166272202295582,
"recall": 0.9190598453517312,
"support": 29841.0
},
"eval_runtime": 4.6671,
"eval_samples_per_second": 17.141,
"eval_steps_per_second": 2.143,
"eval_weighted avg": {
"f1-score": 0.9394463230910989,
"precision": 0.9398464945991887,
"recall": 0.9399819040916859,
"support": 29841.0
},
"step": 567
},
{
"epoch": 8.0,
"eval_B": {
"f1-score": 0.8948545861297539,
"precision": 0.8651766402307137,
"recall": 0.9266409266409267,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9590139880661254,
"precision": 0.9414702069429106,
"recall": 0.9772240219287316,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.8927241122423641,
"precision": 0.9427035531663825,
"recall": 0.8477773847423653,
"support": 8481.0
},
"eval_accuracy": 0.9382393351429241,
"eval_loss": 0.3103167414665222,
"eval_macro avg": {
"f1-score": 0.9155308954794145,
"precision": 0.916450133446669,
"recall": 0.9172141111040079,
"support": 29841.0
},
"eval_runtime": 4.6339,
"eval_samples_per_second": 17.264,
"eval_steps_per_second": 2.158,
"eval_weighted avg": {
"f1-score": 0.9373896838414373,
"precision": 0.9385098450391195,
"recall": 0.9382393351429241,
"support": 29841.0
},
"step": 648
},
{
"epoch": 9.0,
"eval_B": {
"f1-score": 0.9011235955056179,
"precision": 0.8749090909090909,
"recall": 0.9289575289575289,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9615119937082186,
"precision": 0.9485040973670175,
"recall": 0.9748816346872664,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9013722126929674,
"precision": 0.9380339155935229,
"recall": 0.8674684589081476,
"support": 8481.0
},
"eval_accuracy": 0.9423611809255722,
"eval_loss": 0.32883673906326294,
"eval_macro avg": {
"f1-score": 0.9213359339689347,
"precision": 0.9204823679565438,
"recall": 0.9237692075176476,
"support": 29841.0
},
"eval_runtime": 4.6714,
"eval_samples_per_second": 17.126,
"eval_steps_per_second": 2.141,
"eval_weighted avg": {
"f1-score": 0.9417992341337168,
"precision": 0.9423346276781992,
"recall": 0.9423611809255722,
"support": 29841.0
},
"step": 729
},
{
"epoch": 10.0,
"eval_B": {
"f1-score": 0.8935361216730039,
"precision": 0.8801498127340824,
"recall": 0.9073359073359073,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9632332861646931,
"precision": 0.9596359319351009,
"recall": 0.9668577124345876,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9095462405342556,
"precision": 0.9200241254523522,
"recall": 0.8993043273198915,
"support": 8481.0
},
"eval_accuracy": 0.9450755671726819,
"eval_loss": 0.3041447699069977,
"eval_macro avg": {
"f1-score": 0.9221052161239841,
"precision": 0.9199366233738452,
"recall": 0.9244993156967954,
"support": 29841.0
},
"eval_runtime": 4.6532,
"eval_samples_per_second": 17.193,
"eval_steps_per_second": 2.149,
"eval_weighted avg": {
"f1-score": 0.9449504651463465,
"precision": 0.9449285744355027,
"recall": 0.9450755671726819,
"support": 29841.0
},
"step": 810
},
{
"epoch": 11.0,
"eval_B": {
"f1-score": 0.8924122310305775,
"precision": 0.8729689807976366,
"recall": 0.9127413127413128,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9588723786993913,
"precision": 0.9483330083837005,
"recall": 0.9696486419137802,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.8956965718453683,
"precision": 0.9243507715468574,
"recall": 0.8687654757693668,
"support": 8481.0
},
"eval_accuracy": 0.9385074226735028,
"eval_loss": 0.3531416654586792,
"eval_macro avg": {
"f1-score": 0.9156603938584458,
"precision": 0.9152175869093981,
"recall": 0.9170518101414866,
"support": 29841.0
},
"eval_runtime": 4.6508,
"eval_samples_per_second": 17.201,
"eval_steps_per_second": 2.15,
"eval_weighted avg": {
"f1-score": 0.938033267772811,
"precision": 0.9382465579853486,
"recall": 0.9385074226735028,
"support": 29841.0
},
"step": 891
},
{
"epoch": 12.0,
"eval_B": {
"f1-score": 0.8970199924556771,
"precision": 0.8768436578171092,
"recall": 0.9181467181467181,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9595004897159647,
"precision": 0.9431046931407943,
"recall": 0.9764764515325193,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.8953122104873078,
"precision": 0.9400778210116731,
"recall": 0.8546162009197029,
"support": 8481.0
},
"eval_accuracy": 0.9393116852652391,
"eval_loss": 0.4237186312675476,
"eval_macro avg": {
"f1-score": 0.9172775642196499,
"precision": 0.9200087239898588,
"recall": 0.9164131235329801,
"support": 29841.0
},
"eval_runtime": 4.6388,
"eval_samples_per_second": 17.246,
"eval_steps_per_second": 2.156,
"eval_weighted avg": {
"f1-score": 0.9385463313402296,
"precision": 0.9393689288141548,
"recall": 0.9393116852652391,
"support": 29841.0
},
"step": 972
},
{
"epoch": 12.35,
"grad_norm": 0.5223535895347595,
"learning_rate": 1.506172839506173e-05,
"loss": 0.0196,
"step": 1000
},
{
"epoch": 13.0,
"eval_B": {
"f1-score": 0.8891411275066211,
"precision": 0.8716617210682492,
"recall": 0.9073359073359073,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9610692322951795,
"precision": 0.9563736860287223,
"recall": 0.9658111138798904,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9043145233678416,
"precision": 0.918104495747266,
"recall": 0.8909326730338404,
"support": 8481.0
},
"eval_accuracy": 0.9419925605710264,
"eval_loss": 0.4308999180793762,
"eval_macro avg": {
"f1-score": 0.9181749610565474,
"precision": 0.9153799676147458,
"recall": 0.9213598980832126,
"support": 29841.0
},
"eval_runtime": 4.6555,
"eval_samples_per_second": 17.184,
"eval_steps_per_second": 2.148,
"eval_weighted avg": {
"f1-score": 0.9418177466843106,
"precision": 0.9418211242043584,
"recall": 0.9419925605710264,
"support": 29841.0
},
"step": 1053
},
{
"epoch": 14.0,
"eval_B": {
"f1-score": 0.9025601834161253,
"precision": 0.8933434190620272,
"recall": 0.911969111969112,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.960806215404473,
"precision": 0.9570764513895758,
"recall": 0.9645651632195366,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9030873763261413,
"precision": 0.9131011208870676,
"recall": 0.8932908855087843,
"support": 8481.0
},
"eval_accuracy": 0.9420260715123487,
"eval_loss": 0.4198700487613678,
"eval_macro avg": {
"f1-score": 0.9221512583822467,
"precision": 0.9211736637795568,
"recall": 0.9232750535658111,
"support": 29841.0
},
"eval_runtime": 4.6834,
"eval_samples_per_second": 17.082,
"eval_steps_per_second": 2.135,
"eval_weighted avg": {
"f1-score": 0.9418744743217934,
"precision": 0.9418125843993292,
"recall": 0.9420260715123487,
"support": 29841.0
},
"step": 1134
},
{
"epoch": 15.0,
"eval_B": {
"f1-score": 0.8935681470137825,
"precision": 0.8861047835990888,
"recall": 0.9011583011583012,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9633398694432723,
"precision": 0.9595530063291139,
"recall": 0.9671567405930726,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9100768726535964,
"precision": 0.92,
"recall": 0.9003655229336163,
"support": 8481.0
},
"eval_accuracy": 0.9453101437619382,
"eval_loss": 0.39179423451423645,
"eval_macro avg": {
"f1-score": 0.9223282963702171,
"precision": 0.9218859299760677,
"recall": 0.9228935215616634,
"support": 29841.0
},
"eval_runtime": 4.6645,
"eval_samples_per_second": 17.151,
"eval_steps_per_second": 2.144,
"eval_weighted avg": {
"f1-score": 0.945174330208011,
"precision": 0.9451243847979119,
"recall": 0.9453101437619382,
"support": 29841.0
},
"step": 1215
},
{
"epoch": 16.0,
"eval_B": {
"f1-score": 0.9022614028363357,
"precision": 0.895738203957382,
"recall": 0.9088803088803089,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9632045864531594,
"precision": 0.9552494853445741,
"recall": 0.9712932967854473,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9083463808262073,
"precision": 0.9282461538461538,
"recall": 0.8892819243013795,
"support": 8481.0
},
"eval_accuracy": 0.9452766328206159,
"eval_loss": 0.44052162766456604,
"eval_macro avg": {
"f1-score": 0.9246041233719008,
"precision": 0.92641128104937,
"recall": 0.9231518433223785,
"support": 29841.0
},
"eval_runtime": 4.6376,
"eval_samples_per_second": 17.25,
"eval_steps_per_second": 2.156,
"eval_weighted avg": {
"f1-score": 0.9449688080038458,
"precision": 0.944992377210312,
"recall": 0.9452766328206159,
"support": 29841.0
},
"step": 1296
},
{
"epoch": 17.0,
"eval_B": {
"f1-score": 0.8923649906890131,
"precision": 0.8618705035971223,
"recall": 0.9250965250965251,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9599605522682445,
"precision": 0.9498902171261283,
"recall": 0.97024669823075,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9001642635517431,
"precision": 0.9298642533936652,
"recall": 0.8723027944817828,
"support": 8481.0
},
"eval_accuracy": 0.9404510572701987,
"eval_loss": 0.4657233655452728,
"eval_macro avg": {
"f1-score": 0.9174966021696669,
"precision": 0.9138749913723053,
"recall": 0.9225486726030193,
"support": 29841.0
},
"eval_runtime": 4.6396,
"eval_samples_per_second": 17.243,
"eval_steps_per_second": 2.155,
"eval_weighted avg": {
"f1-score": 0.940032648483192,
"precision": 0.9403789498282803,
"recall": 0.9404510572701987,
"support": 29841.0
},
"step": 1377
},
{
"epoch": 18.0,
"eval_B": {
"f1-score": 0.9014302280633939,
"precision": 0.9024767801857585,
"recall": 0.9003861003861003,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9638021281301453,
"precision": 0.9506496024820632,
"recall": 0.9773236979815599,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9083038653822705,
"precision": 0.9404115641964399,
"recall": 0.87831623629289,
"support": 8481.0
},
"eval_accuracy": 0.9458463188230958,
"eval_loss": 0.4731404185295105,
"eval_macro avg": {
"f1-score": 0.9245120738586032,
"precision": 0.9311793156214204,
"recall": 0.9186753448868501,
"support": 29841.0
},
"eval_runtime": 4.6405,
"eval_samples_per_second": 17.24,
"eval_steps_per_second": 2.155,
"eval_weighted avg": {
"f1-score": 0.9453224398840687,
"precision": 0.9456493475450943,
"recall": 0.9458463188230958,
"support": 29841.0
},
"step": 1458
},
{
"epoch": 18.52,
"grad_norm": 1.548362135887146,
"learning_rate": 1.2592592592592593e-05,
"loss": 0.0038,
"step": 1500
},
{
"epoch": 19.0,
"eval_B": {
"f1-score": 0.8891402714932127,
"precision": 0.868828297715549,
"recall": 0.9104247104247104,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9603099696984749,
"precision": 0.9571718572065158,
"recall": 0.9634687266384251,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9027910305343512,
"precision": 0.913358271992277,
"recall": 0.8924655111425539,
"support": 8481.0
},
"eval_accuracy": 0.9409872323313562,
"eval_loss": 0.49517783522605896,
"eval_macro avg": {
"f1-score": 0.9174137572420129,
"precision": 0.9131194756381139,
"recall": 0.9221196494018965,
"support": 29841.0
},
"eval_runtime": 4.6377,
"eval_samples_per_second": 17.25,
"eval_steps_per_second": 2.156,
"eval_weighted avg": {
"f1-score": 0.9408741973642117,
"precision": 0.9408859443435835,
"recall": 0.9409872323313562,
"support": 29841.0
},
"step": 1539
},
{
"epoch": 20.0,
"eval_B": {
"f1-score": 0.899884925201381,
"precision": 0.8940548780487805,
"recall": 0.9057915057915058,
"support": 1295.0
},
"eval_I": {
"f1-score": 0.9634953846153845,
"precision": 0.951896887159533,
"recall": 0.9753800149514079,
"support": 20065.0
},
"eval_O": {
"f1-score": 0.9073556231003039,
"precision": 0.9365039528171665,
"recall": 0.8799669850253508,
"support": 8481.0
},
"eval_accuracy": 0.9452431218792936,
"eval_loss": 0.45907479524612427,
"eval_macro avg": {
"f1-score": 0.9235786443056897,
"precision": 0.9274852393418267,
"recall": 0.9203795019227549,
"support": 29841.0
},
"eval_runtime": 4.6356,
"eval_samples_per_second": 17.258,
"eval_steps_per_second": 2.157,
"eval_weighted avg": {
"f1-score": 0.9447796290324438,
"precision": 0.9450119678218422,
"recall": 0.9452431218792936,
"support": 29841.0
},
"step": 1620
}
],
"logging_steps": 500,
"max_steps": 4050,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"total_flos": 2875733054712000.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}