longformer-sep_tok / checkpoint-492 /trainer_state.json
Theoreticallyhugo's picture
Training in progress, epoch 12, checkpoint
ec99bfd verified
raw
history blame
16.1 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 12.0,
"eval_steps": 500,
"global_step": 492,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_Claim": {
"f1-score": 0.3650793650793651,
"precision": 0.45800144822592326,
"recall": 0.30350287907869483,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.652992277992278,
"precision": 0.6792168674698795,
"recall": 0.6287174721189591,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9929261022378432,
"precision": 0.9994626063591581,
"recall": 0.986474540311174,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.873351637727677,
"precision": 0.8190918322936313,
"recall": 0.9353101963058064,
"support": 12073.0
},
"eval_accuracy": 0.8439319979801381,
"eval_loss": 0.3479689359664917,
"eval_macro avg": {
"f1-score": 0.7210873457592908,
"precision": 0.7389431885871479,
"recall": 0.7135012719536586,
"support": 29705.0
},
"eval_runtime": 1.3927,
"eval_samples_per_second": 57.443,
"eval_steps_per_second": 7.18,
"eval_weighted avg": {
"f1-score": 0.8316056073620907,
"precision": 0.8269800178224755,
"recall": 0.8439319979801381,
"support": 29705.0
},
"step": 41
},
{
"epoch": 2.0,
"eval_Claim": {
"f1-score": 0.427892234548336,
"precision": 0.6302521008403361,
"recall": 0.32389635316698656,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.7983248842847697,
"precision": 0.7593291404612159,
"recall": 0.841542750929368,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9990268931351733,
"precision": 0.9997344192634561,
"recall": 0.9983203677510608,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8914659988441533,
"precision": 0.833381357153148,
"recall": 0.9582539551064359,
"support": 12073.0
},
"eval_accuracy": 0.876047803400101,
"eval_loss": 0.2758464217185974,
"eval_macro avg": {
"f1-score": 0.779177502703108,
"precision": 0.805674254429539,
"recall": 0.7805033567384627,
"support": 29705.0
},
"eval_runtime": 1.3892,
"eval_samples_per_second": 57.586,
"eval_steps_per_second": 7.198,
"eval_weighted avg": {
"f1-score": 0.8606332672536217,
"precision": 0.8628640276786139,
"recall": 0.876047803400101,
"support": 29705.0
},
"step": 82
},
{
"epoch": 3.0,
"eval_Claim": {
"f1-score": 0.5882352941176471,
"precision": 0.620671283963772,
"recall": 0.559021113243762,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.822195416164053,
"precision": 0.8549924736578023,
"recall": 0.79182156133829,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.999557835160948,
"precision": 0.9999115357395613,
"recall": 0.9992043847241867,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.897156953936992,
"precision": 0.8765607712976135,
"recall": 0.918744305475027,
"support": 12073.0
},
"eval_accuracy": 0.8897155361050328,
"eval_loss": 0.2410457581281662,
"eval_macro avg": {
"f1-score": 0.8267863748449099,
"precision": 0.8380340161646872,
"recall": 0.8171978411953165,
"support": 29705.0
},
"eval_runtime": 1.3926,
"eval_samples_per_second": 57.447,
"eval_steps_per_second": 7.181,
"eval_weighted avg": {
"f1-score": 0.8873759763571568,
"precision": 0.8860669651248813,
"recall": 0.8897155361050328,
"support": 29705.0
},
"step": 123
},
{
"epoch": 4.0,
"eval_Claim": {
"f1-score": 0.6018413419094463,
"precision": 0.6344057431534167,
"recall": 0.5724568138195777,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8690502303136654,
"precision": 0.8230162027420025,
"recall": 0.9205390334572491,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9995578742594394,
"precision": 0.9998231027772864,
"recall": 0.9992927864214993,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8916227781435153,
"precision": 0.8858637887335459,
"recall": 0.8974571357574753,
"support": 12073.0
},
"eval_accuracy": 0.8923076923076924,
"eval_loss": 0.24871428310871124,
"eval_macro avg": {
"f1-score": 0.8405180561565165,
"precision": 0.8357772093515627,
"recall": 0.8474364423639503,
"support": 29705.0
},
"eval_runtime": 1.3944,
"eval_samples_per_second": 57.373,
"eval_steps_per_second": 7.172,
"eval_weighted avg": {
"f1-score": 0.8904302737876794,
"precision": 0.8894248936462209,
"recall": 0.8923076923076924,
"support": 29705.0
},
"step": 164
},
{
"epoch": 5.0,
"eval_Claim": {
"f1-score": 0.6216759248315801,
"precision": 0.6126252038201724,
"recall": 0.6309980806142035,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8738404452690167,
"precision": 0.8722222222222222,
"recall": 0.8754646840148699,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9996462681287585,
"precision": 1.0,
"recall": 0.9992927864214993,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8883893260064111,
"precision": 0.8930364914630063,
"recall": 0.8837902758220824,
"support": 12073.0
},
"eval_accuracy": 0.8917017337148628,
"eval_loss": 0.2594112753868103,
"eval_macro avg": {
"f1-score": 0.8458879910589416,
"precision": 0.8444709793763502,
"recall": 0.8473864567181638,
"support": 29705.0
},
"eval_runtime": 1.3912,
"eval_samples_per_second": 57.505,
"eval_steps_per_second": 7.188,
"eval_weighted avg": {
"f1-score": 0.8922798455096741,
"precision": 0.8929161297147813,
"recall": 0.8917017337148628,
"support": 29705.0
},
"step": 205
},
{
"epoch": 6.0,
"eval_Claim": {
"f1-score": 0.6567796610169492,
"precision": 0.5880121396054628,
"recall": 0.7437619961612284,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8874388254486133,
"precision": 0.8901355773726041,
"recall": 0.8847583643122676,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9997789469030461,
"precision": 1.0,
"recall": 0.999557991513437,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8790980052038161,
"precision": 0.9225448257031037,
"recall": 0.8395593473039012,
"support": 12073.0
},
"eval_accuracy": 0.8903214946978623,
"eval_loss": 0.2811674177646637,
"eval_macro avg": {
"f1-score": 0.8557738596431061,
"precision": 0.8501731356702926,
"recall": 0.8669094248227085,
"support": 29705.0
},
"eval_runtime": 1.3897,
"eval_samples_per_second": 57.567,
"eval_steps_per_second": 7.196,
"eval_weighted avg": {
"f1-score": 0.8944647582453118,
"precision": 0.9027534099005212,
"recall": 0.8903214946978623,
"support": 29705.0
},
"step": 246
},
{
"epoch": 7.0,
"eval_Claim": {
"f1-score": 0.6388127853881279,
"precision": 0.6093205574912892,
"recall": 0.6713051823416507,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8804010503700167,
"precision": 0.905252822778596,
"recall": 0.8568773234200744,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9999115904871364,
"precision": 1.0,
"recall": 0.9998231966053748,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.886362682998448,
"precision": 0.8979262281149074,
"recall": 0.8750931831359231,
"support": 12073.0
},
"eval_accuracy": 0.8926780003366437,
"eval_loss": 0.30266591906547546,
"eval_macro avg": {
"f1-score": 0.8513720273109323,
"precision": 0.8531249020961982,
"recall": 0.8507747213757557,
"support": 29705.0
},
"eval_runtime": 1.3903,
"eval_samples_per_second": 57.544,
"eval_steps_per_second": 7.193,
"eval_weighted avg": {
"f1-score": 0.894437008359695,
"precision": 0.8968327052777144,
"recall": 0.8926780003366437,
"support": 29705.0
},
"step": 287
},
{
"epoch": 8.0,
"eval_Claim": {
"f1-score": 0.64190800681431,
"precision": 0.6094457623463446,
"recall": 0.6780230326295585,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8679334916864608,
"precision": 0.8877551020408163,
"recall": 0.8489776951672863,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9996904982977407,
"precision": 1.0,
"recall": 0.9993811881188119,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8887206661619985,
"precision": 0.9026911576249466,
"recall": 0.875176012590077,
"support": 12073.0
},
"eval_accuracy": 0.8929136509005218,
"eval_loss": 0.3308269679546356,
"eval_macro avg": {
"f1-score": 0.8495631657401276,
"precision": 0.8499730055030268,
"recall": 0.8503894821264335,
"support": 29705.0
},
"eval_runtime": 1.3922,
"eval_samples_per_second": 57.462,
"eval_steps_per_second": 7.183,
"eval_weighted avg": {
"f1-score": 0.8948422476293271,
"precision": 0.8975192480409824,
"recall": 0.8929136509005218,
"support": 29705.0
},
"step": 328
},
{
"epoch": 9.0,
"eval_Claim": {
"f1-score": 0.63868836412578,
"precision": 0.651685393258427,
"recall": 0.6261996161228407,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8941736028537456,
"precision": 0.9157330735509012,
"recall": 0.8736059479553904,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9996904982977407,
"precision": 1.0,
"recall": 0.9993811881188119,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8962523039115298,
"precision": 0.8864851725814292,
"recall": 0.9062370578977884,
"support": 12073.0
},
"eval_accuracy": 0.9000504965494025,
"eval_loss": 0.34075838327407837,
"eval_macro avg": {
"f1-score": 0.857201192297199,
"precision": 0.8634759098476893,
"recall": 0.8513559525237079,
"support": 29705.0
},
"eval_runtime": 1.3948,
"eval_samples_per_second": 57.357,
"eval_steps_per_second": 7.17,
"eval_weighted avg": {
"f1-score": 0.8993525560304815,
"precision": 0.8988863080948749,
"recall": 0.9000504965494025,
"support": 29705.0
},
"step": 369
},
{
"epoch": 10.0,
"eval_Claim": {
"f1-score": 0.6206814955040227,
"precision": 0.6122782446311859,
"recall": 0.6293186180422264,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8776916451335055,
"precision": 0.8178170144462279,
"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": 0.8864029718434716,
"precision": 0.9038395316804407,
"recall": 0.869626439161766,
"support": 12073.0
},
"eval_accuracy": 0.89116310385457,
"eval_loss": 0.4050362706184387,
"eval_macro avg": {
"f1-score": 0.8461829783964931,
"precision": 0.8334615992186778,
"recall": 0.8614927698772062,
"support": 29705.0
},
"eval_runtime": 1.3901,
"eval_samples_per_second": 57.551,
"eval_steps_per_second": 7.194,
"eval_weighted avg": {
"f1-score": 0.8917298769484514,
"precision": 0.8932830396594146,
"recall": 0.89116310385457,
"support": 29705.0
},
"step": 410
},
{
"epoch": 11.0,
"eval_Claim": {
"f1-score": 0.6361489146327746,
"precision": 0.5987719669701461,
"recall": 0.6785028790786948,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8838862559241706,
"precision": 0.9018375241779497,
"recall": 0.866635687732342,
"support": 2152.0
},
"eval_O": {
"f1-score": 1.0,
"precision": 1.0,
"recall": 1.0,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8829567053854277,
"precision": 0.90087915876573,
"recall": 0.8657334548165327,
"support": 12073.0
},
"eval_accuracy": 0.8906581383605454,
"eval_loss": 0.4124489724636078,
"eval_macro avg": {
"f1-score": 0.8507479689855932,
"precision": 0.8503721624784565,
"recall": 0.8527180054068924,
"support": 29705.0
},
"eval_runtime": 1.3986,
"eval_samples_per_second": 57.198,
"eval_steps_per_second": 7.15,
"eval_weighted avg": {
"f1-score": 0.8929650968879477,
"precision": 0.8963053356048198,
"recall": 0.8906581383605454,
"support": 29705.0
},
"step": 451
},
{
"epoch": 12.0,
"eval_Claim": {
"f1-score": 0.6082852648138437,
"precision": 0.6705202312138728,
"recall": 0.5566218809980806,
"support": 4168.0
},
"eval_MajorClaim": {
"f1-score": 0.8887310606060606,
"precision": 0.9058880308880309,
"recall": 0.8722118959107806,
"support": 2152.0
},
"eval_O": {
"f1-score": 0.9996462994075515,
"precision": 0.9999115513886432,
"recall": 0.9993811881188119,
"support": 11312.0
},
"eval_Premise": {
"f1-score": 0.8976744186046511,
"precision": 0.8699774617237895,
"recall": 0.9271929097987244,
"support": 12073.0
},
"eval_accuracy": 0.8987039218986702,
"eval_loss": 0.4420628547668457,
"eval_macro avg": {
"f1-score": 0.8485842608580267,
"precision": 0.861574318803584,
"recall": 0.8388519687065994,
"support": 29705.0
},
"eval_runtime": 1.3926,
"eval_samples_per_second": 57.446,
"eval_steps_per_second": 7.181,
"eval_weighted avg": {
"f1-score": 0.8952534731823099,
"precision": 0.8940729416216161,
"recall": 0.8987039218986702,
"support": 29705.0
},
"step": 492
}
],
"logging_steps": 500,
"max_steps": 656,
"num_input_tokens_seen": 0,
"num_train_epochs": 16,
"save_steps": 500,
"total_flos": 1725464792721600.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}