longformer-sep_tok / checkpoint-1134 /trainer_state.json
Theoreticallyhugo's picture
Training in progress, epoch 14, checkpoint
c7a0767 verified
raw
history blame
18.9 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": 0.5254768748366867,
"precision": 0.5768789443488239,
"recall": 0.4824856046065259,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.7984361424847959,
"precision": 0.7495921696574225,
"recall": 0.854089219330855,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9990277532260916,
"precision": 0.9988511841640155,
"recall": 0.9992043847241867,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8854183656825966,
"precision": 0.8719781543651113,
"recall": 0.8992793837488611,
"support": 12073.0
},
"eval_accuracy": 0.8755765022723447,
"eval_loss": 0.28303390741348267,
"eval_macro avg": {
"f1-score": 0.8020897840575427,
"precision": 0.7993251131338432,
"recall": 0.8087646481026072,
"support": 29705.0
},
"eval_runtime": 4.8106,
"eval_samples_per_second": 16.63,
"eval_steps_per_second": 2.079,
"eval_weighted avg": {
"f1-score": 0.8718761173649596,
"precision": 0.8700202202343714,
"recall": 0.8755765022723447,
"support": 29705.0
},
"step": 81
},
{
"epoch": 2.0,
"eval_Claim": {
"f1-score": 0.5578782336033447,
"precision": 0.6124497991967871,
"recall": 0.5122360844529751,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.784421283598464,
"precision": 0.9571619812583668,
"recall": 0.6644981412639405,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9998674150351351,
"precision": 0.9997348652231551,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8977749872463995,
"precision": 0.8530201342281879,
"recall": 0.9474861260664292,
"support": 12073.0
},
"eval_accuracy": 0.8859114627167144,
"eval_loss": 0.27363118529319763,
"eval_macro avg": {
"f1-score": 0.8099854798708358,
"precision": 0.8555916949766242,
"recall": 0.7810550879458362,
"support": 29705.0
},
"eval_runtime": 4.7727,
"eval_samples_per_second": 16.762,
"eval_steps_per_second": 2.095,
"eval_weighted avg": {
"f1-score": 0.8807489883812781,
"precision": 0.8826802296805741,
"recall": 0.8859114627167144,
"support": 29705.0
},
"step": 162
},
{
"epoch": 3.0,
"eval_Claim": {
"f1-score": 0.64760736196319,
"precision": 0.6627322953289804,
"recall": 0.6331573896353166,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9016018306636155,
"precision": 0.8881875563570785,
"recall": 0.9154275092936803,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9996904982977407,
"precision": 1.0,
"recall": 0.9993811881188119,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8982820417748115,
"precision": 0.8936065573770492,
"recall": 0.9030067091857865,
"support": 12073.0
},
"eval_accuracy": 0.9027436458508669,
"eval_loss": 0.22585685551166534,
"eval_macro avg": {
"f1-score": 0.8617954331748394,
"precision": 0.861131602265777,
"recall": 0.8627431990583988,
"support": 29705.0
},
"eval_runtime": 4.8003,
"eval_samples_per_second": 16.666,
"eval_steps_per_second": 2.083,
"eval_weighted avg": {
"f1-score": 0.9019670975035187,
"precision": 0.9013351218793044,
"recall": 0.9027436458508669,
"support": 29705.0
},
"step": 243
},
{
"epoch": 4.0,
"eval_Claim": {
"f1-score": 0.6477602966229081,
"precision": 0.5561865427637239,
"recall": 0.7754318618042226,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8984409799554566,
"precision": 0.8627031650983746,
"recall": 0.9372676579925651,
"support": 2152.0
},
"eval_O": {
"f1-score": 1.0,
"precision": 1.0,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.861495720750997,
"precision": 0.9384029675907849,
"recall": 0.7962395427814131,
"support": 12073.0
},
"eval_accuracy": 0.8811311227066151,
"eval_loss": 0.32341843843460083,
"eval_macro avg": {
"f1-score": 0.8519242493323405,
"precision": 0.8393231688632209,
"recall": 0.8772347656445502,
"support": 29705.0
},
"eval_runtime": 4.7708,
"eval_samples_per_second": 16.769,
"eval_steps_per_second": 2.096,
"eval_weighted avg": {
"f1-score": 0.8869263673393438,
"precision": 0.9027457246003854,
"recall": 0.8811311227066151,
"support": 29705.0
},
"step": 324
},
{
"epoch": 5.0,
"eval_Claim": {
"f1-score": 0.6188788236838649,
"precision": 0.6833864888373442,
"recall": 0.5654990403071017,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.909733679000707,
"precision": 0.9230033476805356,
"recall": 0.8968401486988847,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9999558011049724,
"precision": 0.9999116061168567,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8978936810431294,
"precision": 0.8706816059757236,
"recall": 0.9268615919821088,
"support": 12073.0
},
"eval_accuracy": 0.9018347079616226,
"eval_loss": 0.3072384297847748,
"eval_macro avg": {
"f1-score": 0.8566154962081685,
"precision": 0.8692457621526151,
"recall": 0.8473001952470238,
"support": 29705.0
},
"eval_runtime": 4.7922,
"eval_samples_per_second": 16.694,
"eval_steps_per_second": 2.087,
"eval_weighted avg": {
"f1-score": 0.8984684143294738,
"precision": 0.8974043833368577,
"recall": 0.9018347079616226,
"support": 29705.0
},
"step": 405
},
{
"epoch": 6.0,
"eval_Claim": {
"f1-score": 0.620837537840565,
"precision": 0.6545212765957447,
"recall": 0.5904510556621881,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8993613741466637,
"precision": 0.8547509418166597,
"recall": 0.9488847583643123,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9999558011049724,
"precision": 0.9999116061168567,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8931567692054615,
"precision": 0.8869558114841134,
"recall": 0.8994450426571688,
"support": 12073.0
},
"eval_accuracy": 0.8979633058407676,
"eval_loss": 0.3634909987449646,
"eval_macro avg": {
"f1-score": 0.8533278705744156,
"precision": 0.8490349090033436,
"recall": 0.8596952141709173,
"support": 29705.0
},
"eval_runtime": 4.7765,
"eval_samples_per_second": 16.749,
"eval_steps_per_second": 2.094,
"eval_weighted avg": {
"f1-score": 0.8960665959131486,
"precision": 0.8950239457358053,
"recall": 0.8979633058407676,
"support": 29705.0
},
"step": 486
},
{
"epoch": 6.17,
"grad_norm": 9.48759651184082,
"learning_rate": 1.7530864197530865e-05,
"loss": 0.214,
"step": 500
},
{
"epoch": 7.0,
"eval_Claim": {
"f1-score": 0.6812872467222885,
"precision": 0.6769303647560397,
"recall": 0.685700575815739,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9276377217553688,
"precision": 0.9319887429643527,
"recall": 0.9233271375464684,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9999115983026874,
"precision": 0.9999115983026874,
"recall": 0.9999115983026874,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.9017086927670869,
"precision": 0.9029819752471135,
"recall": 0.9004389961070156,
"support": 12073.0
},
"eval_accuracy": 0.9098468271334792,
"eval_loss": 0.3468942642211914,
"eval_macro avg": {
"f1-score": 0.8776363148868579,
"precision": 0.8779531703175483,
"recall": 0.8773445769429775,
"support": 29705.0
},
"eval_runtime": 4.7913,
"eval_samples_per_second": 16.697,
"eval_steps_per_second": 2.087,
"eval_weighted avg": {
"f1-score": 0.9100559053806798,
"precision": 0.910277290769933,
"recall": 0.9098468271334792,
"support": 29705.0
},
"step": 567
},
{
"epoch": 8.0,
"eval_Claim": {
"f1-score": 0.6928905033731187,
"precision": 0.6105725260654838,
"recall": 0.8008637236084453,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9302433371958285,
"precision": 0.9278779472954231,
"recall": 0.9326208178438662,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9999115904871364,
"precision": 1.0,
"recall": 0.9998231966053748,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8820387074174622,
"precision": 0.9356247097073851,
"recall": 0.8342582622380519,
"support": 12073.0
},
"eval_accuracy": 0.8997475172529877,
"eval_loss": 0.40174800157546997,
"eval_macro avg": {
"f1-score": 0.8762710346183864,
"precision": 0.868518795767073,
"recall": 0.8918915000739345,
"support": 29705.0
},
"eval_runtime": 4.7722,
"eval_samples_per_second": 16.764,
"eval_steps_per_second": 2.095,
"eval_weighted avg": {
"f1-score": 0.9038782866839282,
"precision": 0.9139692560686062,
"recall": 0.8997475172529877,
"support": 29705.0
},
"step": 648
},
{
"epoch": 9.0,
"eval_Claim": {
"f1-score": 0.6602241979126401,
"precision": 0.7130531589201224,
"recall": 0.6146833013435701,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9052154195011338,
"precision": 0.8839681133746679,
"recall": 0.9275092936802974,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9999558011049724,
"precision": 0.9999116061168567,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.9072885349800925,
"precision": 0.8903596204449405,
"recall": 0.9248736850824153,
"support": 12073.0
},
"eval_accuracy": 0.910149806429894,
"eval_loss": 0.4713464379310608,
"eval_macro avg": {
"f1-score": 0.8681709883747097,
"precision": 0.8718231247141469,
"recall": 0.8667665700265708,
"support": 29705.0
},
"eval_runtime": 4.7731,
"eval_samples_per_second": 16.761,
"eval_steps_per_second": 2.095,
"eval_weighted avg": {
"f1-score": 0.9077607320175198,
"precision": 0.9067368029754924,
"recall": 0.910149806429894,
"support": 29705.0
},
"step": 729
},
{
"epoch": 10.0,
"eval_Claim": {
"f1-score": 0.6952499153785399,
"precision": 0.6562300319488817,
"recall": 0.7392034548944337,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9103908484270734,
"precision": 0.9344422700587084,
"recall": 0.887546468401487,
"support": 2152.0
},
"eval_O": {
"f1-score": 1.0,
"precision": 1.0,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.9012517385257303,
"precision": 0.9174532349407929,
"recall": 0.8856125238134681,
"support": 12073.0
},
"eval_accuracy": 0.9087695674128935,
"eval_loss": 0.4968256950378418,
"eval_macro avg": {
"f1-score": 0.8767231255828358,
"precision": 0.8770313842370958,
"recall": 0.8780906117773473,
"support": 29705.0
},
"eval_runtime": 4.7941,
"eval_samples_per_second": 16.687,
"eval_steps_per_second": 2.086,
"eval_weighted avg": {
"f1-score": 0.9106135328171674,
"precision": 0.9134657277821738,
"recall": 0.9087695674128935,
"support": 29705.0
},
"step": 810
},
{
"epoch": 11.0,
"eval_Claim": {
"f1-score": 0.6796779702366431,
"precision": 0.6913151364764268,
"recall": 0.668426103646833,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9301903898458749,
"precision": 0.9079646017699115,
"recall": 0.9535315985130112,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9999558011049724,
"precision": 0.9999116061168567,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.901923474663909,
"precision": 0.9008428358948934,
"recall": 0.9030067091857865,
"support": 12073.0
},
"eval_accuracy": 0.9106884362901868,
"eval_loss": 0.5864725708961487,
"eval_macro avg": {
"f1-score": 0.8779369089628499,
"precision": 0.8750085450645222,
"recall": 0.8812411028364077,
"support": 29705.0
},
"eval_runtime": 4.8031,
"eval_samples_per_second": 16.656,
"eval_steps_per_second": 2.082,
"eval_weighted avg": {
"f1-score": 0.9101191594213591,
"precision": 0.9096858090555641,
"recall": 0.9106884362901868,
"support": 29705.0
},
"step": 891
},
{
"epoch": 12.0,
"eval_Claim": {
"f1-score": 0.697112745636184,
"precision": 0.6425824731835661,
"recall": 0.7617562380038387,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9365296803652968,
"precision": 0.920556552962298,
"recall": 0.9530669144981413,
"support": 2152.0
},
"eval_O": {
"f1-score": 1.0,
"precision": 1.0,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8923037300940034,
"precision": 0.9260513186029936,
"recall": 0.8609293464756067,
"support": 12073.0
},
"eval_accuracy": 0.9066487123379903,
"eval_loss": 0.5556066632270813,
"eval_macro avg": {
"f1-score": 0.8814865390238711,
"precision": 0.8722975861872144,
"recall": 0.8939381247443967,
"support": 29705.0
},
"eval_runtime": 4.7999,
"eval_samples_per_second": 16.667,
"eval_steps_per_second": 2.083,
"eval_weighted avg": {
"f1-score": 0.9091318205481447,
"precision": 0.9140393543072853,
"recall": 0.9066487123379903,
"support": 29705.0
},
"step": 972
},
{
"epoch": 12.35,
"grad_norm": 3.987734794616699,
"learning_rate": 1.506172839506173e-05,
"loss": 0.0252,
"step": 1000
},
{
"epoch": 13.0,
"eval_Claim": {
"f1-score": 0.6895428929242329,
"precision": 0.7212470526591565,
"recall": 0.6605086372360844,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9265742214139577,
"precision": 0.9069870939029817,
"recall": 0.9470260223048327,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9999558011049724,
"precision": 0.9999116061168567,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.9083234293676489,
"precision": 0.8989292667099286,
"recall": 0.917916010933488,
"support": 12073.0
},
"eval_accuracy": 0.9151657970038714,
"eval_loss": 0.5547633767127991,
"eval_macro avg": {
"f1-score": 0.8810990862027029,
"precision": 0.8817687548472308,
"recall": 0.8813626676186013,
"support": 29705.0
},
"eval_runtime": 4.8002,
"eval_samples_per_second": 16.666,
"eval_steps_per_second": 2.083,
"eval_weighted avg": {
"f1-score": 0.9138424940934561,
"precision": 0.9130371003853032,
"recall": 0.9151657970038714,
"support": 29705.0
},
"step": 1053
},
{
"epoch": 14.0,
"eval_Claim": {
"f1-score": 0.6662641873943491,
"precision": 0.6706368497812348,
"recall": 0.6619481765834933,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.9252051582649472,
"precision": 0.9337434926644581,
"recall": 0.91682156133829,
"support": 2152.0
},
"eval_O": {
"f1-score": 1.0,
"precision": 1.0,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8968191757085687,
"precision": 0.8933914187078744,
"recall": 0.9002733371987078,
"support": 12073.0
},
"eval_accuracy": 0.9060090893788925,
"eval_loss": 0.598924994468689,
"eval_macro avg": {
"f1-score": 0.8720721303419663,
"precision": 0.8744429402883919,
"recall": 0.8697607687801228,
"support": 29705.0
},
"eval_runtime": 4.8166,
"eval_samples_per_second": 16.609,
"eval_steps_per_second": 2.076,
"eval_weighted avg": {
"f1-score": 0.9058181633386757,
"precision": 0.9056571278963901,
"recall": 0.9060090893788925,
"support": 29705.0
},
"step": 1134
}
],
"logging_steps": 500,
"max_steps": 4050,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"total_flos": 2013042258175200.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}