longformer-sep_tok / checkpoint-1296 /trainer_state.json
Theoreticallyhugo's picture
Training in progress, epoch 16, checkpoint
c393310 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": 0.5724454649827784,
"precision": 0.6002889477486154,
"recall": 0.5470704410796576,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.7860447185813415,
"precision": 0.698526892771497,
"recall": 0.8986337593653592,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9987820830889982,
"precision": 0.9975671292124707,
"recall": 1.0,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.8965445389824622,
"precision": 0.9051258677512096,
"recall": 0.8881243979633962,
"support": 14534.0
},
"eval_accuracy": 0.8791279948197712,
"eval_loss": 0.3157936632633209,
"eval_macro avg": {
"f1-score": 0.8134542014088951,
"precision": 0.8003772093709481,
"recall": 0.8334571496021033,
"support": 32431.0
},
"eval_runtime": 4.8184,
"eval_samples_per_second": 16.603,
"eval_steps_per_second": 2.075,
"eval_weighted avg": {
"f1-score": 0.8781740995293698,
"precision": 0.8793943851252293,
"recall": 0.8791279948197712,
"support": 32431.0
},
"step": 81
},
{
"epoch": 2.0,
"eval_Claim": {
"f1-score": 0.6090644692195831,
"precision": 0.680108254397835,
"recall": 0.5514592933947773,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8585858585858586,
"precision": 0.8960229995208433,
"recall": 0.8241516086381666,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.999458141425088,
"precision": 0.9992776523702032,
"recall": 0.9996386956914461,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9105885478942475,
"precision": 0.8801849235905997,
"recall": 0.9431677445988716,
"support": 14534.0
},
"eval_accuracy": 0.8990780426135487,
"eval_loss": 0.24451124668121338,
"eval_macro avg": {
"f1-score": 0.8444242542811943,
"precision": 0.8638984574698704,
"recall": 0.8296043355808155,
"support": 32431.0
},
"eval_runtime": 4.8529,
"eval_samples_per_second": 16.485,
"eval_steps_per_second": 2.061,
"eval_weighted avg": {
"f1-score": 0.8949194640367888,
"precision": 0.8938342965082798,
"recall": 0.8990780426135487,
"support": 32431.0
},
"step": 162
},
{
"epoch": 3.0,
"eval_Claim": {
"f1-score": 0.6542320859402667,
"precision": 0.7079713847726111,
"recall": 0.60807548825982,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8915187376725839,
"precision": 0.8866608544027899,
"recall": 0.8964301454385192,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9995936611133684,
"precision": 0.9992778479870013,
"recall": 0.9999096739228616,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9157990498332153,
"precision": 0.8973258501155497,
"recall": 0.935048850970139,
"support": 14534.0
},
"eval_accuracy": 0.9085442940396534,
"eval_loss": 0.22665414214134216,
"eval_macro avg": {
"f1-score": 0.8652858836398586,
"precision": 0.872808984319488,
"recall": 0.8598660396478349,
"support": 32431.0
},
"eval_runtime": 4.8494,
"eval_samples_per_second": 16.497,
"eval_steps_per_second": 2.062,
"eval_weighted avg": {
"f1-score": 0.9059516032151627,
"precision": 0.9047762338408377,
"recall": 0.9085442940396534,
"support": 32431.0
},
"step": 243
},
{
"epoch": 4.0,
"eval_Claim": {
"f1-score": 0.6861642294713161,
"precision": 0.6424741478360781,
"recall": 0.7362299758613122,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8995515695067264,
"precision": 0.9155636695572797,
"recall": 0.884089907448215,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9976481230212574,
"precision": 0.9990941208442794,
"recall": 0.9962063047601842,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9062532879739067,
"precision": 0.9242435081193219,
"recall": 0.8889500481629283,
"support": 14534.0
},
"eval_accuracy": 0.9037649162838025,
"eval_loss": 0.2452327460050583,
"eval_macro avg": {
"f1-score": 0.8724043024933017,
"precision": 0.8703438615892398,
"recall": 0.8763690590581599,
"support": 32431.0
},
"eval_runtime": 4.8393,
"eval_samples_per_second": 16.531,
"eval_steps_per_second": 2.066,
"eval_weighted avg": {
"f1-score": 0.9060584182508294,
"precision": 0.9095955973231696,
"recall": 0.9037649162838025,
"support": 32431.0
},
"step": 324
},
{
"epoch": 5.0,
"eval_Claim": {
"f1-score": 0.6640249759846301,
"precision": 0.7332272606735614,
"recall": 0.6067588325652842,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8848007008322383,
"precision": 0.8794079233783195,
"recall": 0.8902600264433671,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.999367774566474,
"precision": 0.9992775219001174,
"recall": 0.9994580435371692,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9201314377682404,
"precision": 0.8973839110529758,
"recall": 0.9440621989816981,
"support": 14534.0
},
"eval_accuracy": 0.9118127717307515,
"eval_loss": 0.28411003947257996,
"eval_macro avg": {
"f1-score": 0.8670812222878956,
"precision": 0.8773241542512435,
"recall": 0.8601347753818797,
"support": 32431.0
},
"eval_runtime": 4.8228,
"eval_samples_per_second": 16.588,
"eval_steps_per_second": 2.073,
"eval_weighted avg": {
"f1-score": 0.9087220731552942,
"precision": 0.9078434958291443,
"recall": 0.9118127717307515,
"support": 32431.0
},
"step": 405
},
{
"epoch": 6.0,
"eval_Claim": {
"f1-score": 0.6496087442553721,
"precision": 0.7484258729250143,
"recall": 0.5738424402018872,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8865800865800866,
"precision": 0.8711186729051468,
"recall": 0.9026002644336713,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9986892655367232,
"precision": 0.9994572100597069,
"recall": 0.9979225002258152,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9191778088205947,
"precision": 0.889647180015452,
"recall": 0.9507362047612495,
"support": 14534.0
},
"eval_accuracy": 0.9105177145323918,
"eval_loss": 0.3082066774368286,
"eval_macro avg": {
"f1-score": 0.8635139762981943,
"precision": 0.87716223397633,
"recall": 0.8562753524056559,
"support": 32431.0
},
"eval_runtime": 4.8389,
"eval_samples_per_second": 16.533,
"eval_steps_per_second": 2.067,
"eval_weighted avg": {
"f1-score": 0.9061618943658084,
"precision": 0.9059932736781678,
"recall": 0.9105177145323918,
"support": 32431.0
},
"step": 486
},
{
"epoch": 6.17,
"grad_norm": 0.8612700700759888,
"learning_rate": 1.7530864197530865e-05,
"loss": 0.2223,
"step": 500
},
{
"epoch": 7.0,
"eval_Claim": {
"f1-score": 0.6977381768334475,
"precision": 0.629950495049505,
"recall": 0.7818740399385561,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8892209178228387,
"precision": 0.8621688741721855,
"recall": 0.9180255619215514,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9988698521766648,
"precision": 0.9998190045248869,
"recall": 0.9979225002258152,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9034227633516503,
"precision": 0.9449996243143738,
"recall": 0.8653502132929682,
"support": 14534.0
},
"eval_accuracy": 0.9025623631710401,
"eval_loss": 0.3512111008167267,
"eval_macro avg": {
"f1-score": 0.8723129275461503,
"precision": 0.8592344995152377,
"recall": 0.8907930788447227,
"support": 32431.0
},
"eval_runtime": 4.8174,
"eval_samples_per_second": 16.607,
"eval_steps_per_second": 2.076,
"eval_weighted avg": {
"f1-score": 0.9061105026169647,
"precision": 0.9136494810618672,
"recall": 0.9025623631710401,
"support": 32431.0
},
"step": 567
},
{
"epoch": 8.0,
"eval_Claim": {
"f1-score": 0.6987032799389779,
"precision": 0.6177710335525206,
"recall": 0.80403774412991,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.9015364639688379,
"precision": 0.8856292517006803,
"recall": 0.9180255619215514,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.998915499322187,
"precision": 0.9994574554661362,
"recall": 0.9983741306115076,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.8976577489773016,
"precision": 0.9472075788830315,
"recall": 0.8530342644832806,
"support": 14534.0
},
"eval_accuracy": 0.9003114304215103,
"eval_loss": 0.3705728054046631,
"eval_macro avg": {
"f1-score": 0.8742032480518261,
"precision": 0.8625163299005921,
"recall": 0.8933679252865624,
"support": 32431.0
},
"eval_runtime": 4.8542,
"eval_samples_per_second": 16.48,
"eval_steps_per_second": 2.06,
"eval_weighted avg": {
"f1-score": 0.9045397397569994,
"precision": 0.9144455555782818,
"recall": 0.9003114304215103,
"support": 32431.0
},
"step": 648
},
{
"epoch": 9.0,
"eval_Claim": {
"f1-score": 0.6657139454567107,
"precision": 0.7278645833333334,
"recall": 0.6133421110379635,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8882521489971346,
"precision": 0.8884479717813051,
"recall": 0.8880564125165271,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9987339482727438,
"precision": 0.9999094612947035,
"recall": 0.9975611959172613,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9178854152690192,
"precision": 0.8955360649299646,
"recall": 0.9413788358332187,
"support": 14534.0
},
"eval_accuracy": 0.9107335573987851,
"eval_loss": 0.41826051473617554,
"eval_macro avg": {
"f1-score": 0.8676463644989021,
"precision": 0.8779395203348266,
"recall": 0.8600846388262426,
"support": 32431.0
},
"eval_runtime": 4.809,
"eval_samples_per_second": 16.635,
"eval_steps_per_second": 2.079,
"eval_weighted avg": {
"f1-score": 0.9079779452489347,
"precision": 0.9071100542045434,
"recall": 0.9107335573987851,
"support": 32431.0
},
"step": 729
},
{
"epoch": 10.0,
"eval_Claim": {
"f1-score": 0.6738416222952305,
"precision": 0.6964871194379391,
"recall": 0.652622339258284,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8951019550136641,
"precision": 0.8557073954983923,
"recall": 0.9382988100484795,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9989605459393501,
"precision": 0.9996382054992764,
"recall": 0.998283804534369,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9109121470961545,
"precision": 0.9083259218717931,
"recall": 0.9135131415990092,
"support": 14534.0
},
"eval_accuracy": 0.9075267490980852,
"eval_loss": 0.5459077954292297,
"eval_macro avg": {
"f1-score": 0.8697040675860997,
"precision": 0.8650396605768502,
"recall": 0.8756795238600354,
"support": 32431.0
},
"eval_runtime": 4.8478,
"eval_samples_per_second": 16.502,
"eval_steps_per_second": 2.063,
"eval_weighted avg": {
"f1-score": 0.9065515080853328,
"precision": 0.9060496255197702,
"recall": 0.9075267490980852,
"support": 32431.0
},
"step": 810
},
{
"epoch": 11.0,
"eval_Claim": {
"f1-score": 0.6581417175035196,
"precision": 0.7070834383665239,
"recall": 0.6155365371955234,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.881979417560762,
"precision": 0.876414273281114,
"recall": 0.8876156897311591,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9999548390010388,
"precision": 0.9999096820809249,
"recall": 1.0,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9173079519373565,
"precision": 0.9002915065588976,
"recall": 0.9349800467868447,
"support": 14534.0
},
"eval_accuracy": 0.9089759797724399,
"eval_loss": 0.5726844668388367,
"eval_macro avg": {
"f1-score": 0.8643459815006692,
"precision": 0.8709247250718651,
"recall": 0.8595330684283817,
"support": 32431.0
},
"eval_runtime": 4.8447,
"eval_samples_per_second": 16.513,
"eval_steps_per_second": 2.064,
"eval_weighted avg": {
"f1-score": 0.9066330640790278,
"precision": 0.9054793272287636,
"recall": 0.9089759797724399,
"support": 32431.0
},
"step": 891
},
{
"epoch": 12.0,
"eval_Claim": {
"f1-score": 0.7079179557694512,
"precision": 0.7407673860911271,
"recall": 0.6778582400702217,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.8947480331703169,
"precision": 0.8644207066557108,
"recall": 0.9272807404142794,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.999186477447347,
"precision": 0.9999095431931253,
"recall": 0.998464456688646,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9243840851702722,
"precision": 0.9169374492282697,
"recall": 0.9319526627218935,
"support": 14534.0
},
"eval_accuracy": 0.9186272393697388,
"eval_loss": 0.5660926699638367,
"eval_macro avg": {
"f1-score": 0.8815591378893468,
"precision": 0.8805087712920582,
"recall": 0.8838890249737601,
"support": 32431.0
},
"eval_runtime": 4.8659,
"eval_samples_per_second": 16.441,
"eval_steps_per_second": 2.055,
"eval_weighted avg": {
"f1-score": 0.917429533390557,
"precision": 0.9168331226787283,
"recall": 0.9186272393697388,
"support": 32431.0
},
"step": 972
},
{
"epoch": 12.35,
"grad_norm": 2.9101791381835938,
"learning_rate": 1.506172839506173e-05,
"loss": 0.0294,
"step": 1000
},
{
"epoch": 13.0,
"eval_Claim": {
"f1-score": 0.6744648496900223,
"precision": 0.7221943887775552,
"recall": 0.6326530612244898,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.9044719714465672,
"precision": 0.863672814755413,
"recall": 0.9493168796826796,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9995934041111363,
"precision": 0.9999096167751266,
"recall": 0.9992773913828923,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9161992180860105,
"precision": 0.9055171023452725,
"recall": 0.9271363698912893,
"support": 14534.0
},
"eval_accuracy": 0.9119361105115475,
"eval_loss": 0.5764271020889282,
"eval_macro avg": {
"f1-score": 0.8736823608334341,
"precision": 0.8728234806633418,
"recall": 0.8770959255453377,
"support": 32431.0
},
"eval_runtime": 4.8358,
"eval_samples_per_second": 16.543,
"eval_steps_per_second": 2.068,
"eval_weighted avg": {
"f1-score": 0.9098800911419986,
"precision": 0.909053016531805,
"recall": 0.9119361105115475,
"support": 32431.0
},
"step": 1053
},
{
"epoch": 14.0,
"eval_Claim": {
"f1-score": 0.7047413793103449,
"precision": 0.6923565530383231,
"recall": 0.717577353522054,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.9112001687407719,
"precision": 0.8737864077669902,
"recall": 0.9519612163948876,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9992316384180792,
"precision": 1.0,
"recall": 0.998464456688646,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9145424153781864,
"precision": 0.9258919757439007,
"recall": 0.903467730838035,
"support": 14534.0
},
"eval_accuracy": 0.9131694983195091,
"eval_loss": 0.5468704700469971,
"eval_macro avg": {
"f1-score": 0.8824289004618456,
"precision": 0.8730087341373035,
"recall": 0.8928676893609057,
"support": 32431.0
},
"eval_runtime": 4.8411,
"eval_samples_per_second": 16.525,
"eval_steps_per_second": 2.066,
"eval_weighted avg": {
"f1-score": 0.9137390947680665,
"precision": 0.9147298617643856,
"recall": 0.9131694983195091,
"support": 32431.0
},
"step": 1134
},
{
"epoch": 15.0,
"eval_Claim": {
"f1-score": 0.6868638486408813,
"precision": 0.7559304164470216,
"recall": 0.6293614219881501,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.9058546000428909,
"precision": 0.8822055137844611,
"recall": 0.9308065226972234,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9999548349216385,
"precision": 1.0,
"recall": 0.9999096739228616,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9218029420675262,
"precision": 0.9023924075660713,
"recall": 0.9420668776661622,
"support": 14534.0
},
"eval_accuracy": 0.9170855046097869,
"eval_loss": 0.6552095413208008,
"eval_macro avg": {
"f1-score": 0.8786190564182342,
"precision": 0.8851320844493885,
"recall": 0.8755361240685993,
"support": 32431.0
},
"eval_runtime": 4.8081,
"eval_samples_per_second": 16.639,
"eval_steps_per_second": 2.08,
"eval_weighted avg": {
"f1-score": 0.9143537536055225,
"precision": 0.9137205288178378,
"recall": 0.9170855046097869,
"support": 32431.0
},
"step": 1215
},
{
"epoch": 16.0,
"eval_Claim": {
"f1-score": 0.7261954261954263,
"precision": 0.6899071696622556,
"recall": 0.7665130568356375,
"support": 4557.0
},
"eval_MajorClaim": {
"f1-score": 0.9104052573932091,
"precision": 0.9050522648083623,
"recall": 0.9158219479947113,
"support": 2269.0
},
"eval_O": {
"f1-score": 0.9999548349216385,
"precision": 1.0,
"recall": 0.9999096739228616,
"support": 11071.0
},
"eval_Premise": {
"f1-score": 0.9186291000841041,
"precision": 0.9360805599200114,
"recall": 0.9018164304389706,
"support": 14534.0
},
"eval_accuracy": 0.9172705127809812,
"eval_loss": 0.604762613773346,
"eval_macro avg": {
"f1-score": 0.8887961546485946,
"precision": 0.8827599985976573,
"recall": 0.8960152772980452,
"support": 32431.0
},
"eval_runtime": 4.8523,
"eval_samples_per_second": 16.487,
"eval_steps_per_second": 2.061,
"eval_weighted avg": {
"f1-score": 0.918776399254959,
"precision": 0.92113920073012,
"recall": 0.9172705127809812,
"support": 32431.0
},
"step": 1296
}
],
"logging_steps": 500,
"max_steps": 4050,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"total_flos": 2300619723628800.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}