nyt-ingredient-tagger-gte-base / trainer_state.json
napsternxg's picture
End of training
7b4df28
raw
history blame contribute delete
No virus
37.6 kB
{
"best_metric": 0.8753200769424438,
"best_model_checkpoint": "nyt-ingredient-tagger-gte-base/checkpoint-10000",
"epoch": 5.0,
"eval_steps": 1000,
"global_step": 25255,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.1,
"learning_rate": 4.9010097010492975e-05,
"loss": 1.0499,
"step": 500
},
{
"epoch": 0.2,
"learning_rate": 4.802019402098595e-05,
"loss": 0.944,
"step": 1000
},
{
"epoch": 0.2,
"eval_COMMENT": {
"f1": 0.675847596563709,
"number": 6830,
"precision": 0.6120679415607554,
"recall": 0.7544655929721815
},
"eval_NAME": {
"f1": 0.8001553915311616,
"number": 8829,
"precision": 0.7844396082698586,
"recall": 0.8165137614678899
},
"eval_QTY": {
"f1": 0.9824342150940777,
"number": 7140,
"precision": 0.9741153793198403,
"recall": 0.9908963585434174
},
"eval_RANGE_END": {
"f1": 0.7435897435897435,
"number": 94,
"precision": 0.6214285714285714,
"recall": 0.925531914893617
},
"eval_UNIT": {
"f1": 0.9482613808163951,
"number": 5724,
"precision": 0.9131349077968295,
"recall": 0.9861984626135569
},
"eval_loss": 0.9201706647872925,
"eval_overall_accuracy": 0.8220316421997413,
"eval_overall_f1": 0.8416177626189163,
"eval_overall_precision": 0.8068538821568251,
"eval_overall_recall": 0.8795121780759688,
"eval_runtime": 20.4697,
"eval_samples_per_second": 415.541,
"eval_steps_per_second": 12.995,
"step": 1000
},
{
"epoch": 0.3,
"learning_rate": 4.703029103147892e-05,
"loss": 0.9166,
"step": 1500
},
{
"epoch": 0.4,
"learning_rate": 4.6040388041971886e-05,
"loss": 0.9154,
"step": 2000
},
{
"epoch": 0.4,
"eval_COMMENT": {
"f1": 0.7058436863209098,
"number": 6830,
"precision": 0.6404628414648694,
"recall": 0.7860907759882869
},
"eval_NAME": {
"f1": 0.8065916935753257,
"number": 8829,
"precision": 0.7931676338552502,
"recall": 0.8204779703250651
},
"eval_QTY": {
"f1": 0.9844718334377829,
"number": 7140,
"precision": 0.9789502838941975,
"recall": 0.9900560224089636
},
"eval_RANGE_END": {
"f1": 0.7368421052631579,
"number": 94,
"precision": 0.5947712418300654,
"recall": 0.9680851063829787
},
"eval_UNIT": {
"f1": 0.9523889778377012,
"number": 5724,
"precision": 0.9199088393293179,
"recall": 0.9872466806429071
},
"eval_loss": 0.9013546705245972,
"eval_overall_accuracy": 0.8305887274663422,
"eval_overall_f1": 0.8524392288348701,
"eval_overall_precision": 0.819256920052847,
"eval_overall_recall": 0.8884229653702345,
"eval_runtime": 19.4822,
"eval_samples_per_second": 436.604,
"eval_steps_per_second": 13.654,
"step": 2000
},
{
"epoch": 0.49,
"learning_rate": 4.505048505246486e-05,
"loss": 0.9079,
"step": 2500
},
{
"epoch": 0.59,
"learning_rate": 4.406058206295783e-05,
"loss": 0.9086,
"step": 3000
},
{
"epoch": 0.59,
"eval_COMMENT": {
"f1": 0.7271422045291392,
"number": 6830,
"precision": 0.6559519604380077,
"recall": 0.8156661786237189
},
"eval_NAME": {
"f1": 0.8080200501253132,
"number": 8829,
"precision": 0.7948717948717948,
"recall": 0.8216106014271152
},
"eval_QTY": {
"f1": 0.9833923980265444,
"number": 7140,
"precision": 0.975865397876155,
"recall": 0.9910364145658264
},
"eval_RANGE_END": {
"f1": 0.7586206896551724,
"number": 94,
"precision": 0.6376811594202898,
"recall": 0.9361702127659575
},
"eval_UNIT": {
"f1": 0.9514235025766664,
"number": 5724,
"precision": 0.9211516440372975,
"recall": 0.9837526205450734
},
"eval_loss": 0.8954105973243713,
"eval_overall_accuracy": 0.8332889632615806,
"eval_overall_f1": 0.8577454886337004,
"eval_overall_precision": 0.8232383278172295,
"eval_overall_recall": 0.8952720410944544,
"eval_runtime": 20.4332,
"eval_samples_per_second": 416.282,
"eval_steps_per_second": 13.018,
"step": 3000
},
{
"epoch": 0.69,
"learning_rate": 4.3070679073450804e-05,
"loss": 0.8971,
"step": 3500
},
{
"epoch": 0.79,
"learning_rate": 4.208077608394378e-05,
"loss": 0.8962,
"step": 4000
},
{
"epoch": 0.79,
"eval_COMMENT": {
"f1": 0.7278303773836512,
"number": 6830,
"precision": 0.6682174338883448,
"recall": 0.799121522693997
},
"eval_NAME": {
"f1": 0.8103381050520805,
"number": 8829,
"precision": 0.7972380534853135,
"recall": 0.8238758636312153
},
"eval_QTY": {
"f1": 0.984960311934271,
"number": 7140,
"precision": 0.9793685959567987,
"recall": 0.9906162464985995
},
"eval_RANGE_END": {
"f1": 0.7477477477477479,
"number": 94,
"precision": 0.6484375,
"recall": 0.8829787234042553
},
"eval_UNIT": {
"f1": 0.9535336436732967,
"number": 5724,
"precision": 0.9248070924314562,
"recall": 0.9841020265548568
},
"eval_loss": 0.8914348483085632,
"eval_overall_accuracy": 0.8360272305468929,
"eval_overall_f1": 0.860016849199663,
"eval_overall_precision": 0.8304103081378323,
"eval_overall_recall": 0.8918125589684454,
"eval_runtime": 19.856,
"eval_samples_per_second": 428.385,
"eval_steps_per_second": 13.396,
"step": 4000
},
{
"epoch": 0.89,
"learning_rate": 4.109087309443675e-05,
"loss": 0.8884,
"step": 4500
},
{
"epoch": 0.99,
"learning_rate": 4.010097010492972e-05,
"loss": 0.8891,
"step": 5000
},
{
"epoch": 0.99,
"eval_COMMENT": {
"f1": 0.7349575986000808,
"number": 6830,
"precision": 0.680119581464873,
"recall": 0.7994143484626647
},
"eval_NAME": {
"f1": 0.8115361060074606,
"number": 8829,
"precision": 0.7980727113447219,
"recall": 0.8254615471740854
},
"eval_QTY": {
"f1": 0.9848980444011414,
"number": 7140,
"precision": 0.9788352469221192,
"recall": 0.9910364145658264
},
"eval_RANGE_END": {
"f1": 0.7678571428571428,
"number": 94,
"precision": 0.6615384615384615,
"recall": 0.9148936170212766
},
"eval_UNIT": {
"f1": 0.9549260042283298,
"number": 5724,
"precision": 0.9254220619570562,
"recall": 0.9863731656184487
},
"eval_loss": 0.8838098049163818,
"eval_overall_accuracy": 0.8388605765573895,
"eval_overall_f1": 0.8628391039384169,
"eval_overall_precision": 0.8346178967994775,
"eval_overall_recall": 0.8930356082049131,
"eval_runtime": 20.0124,
"eval_samples_per_second": 425.036,
"eval_steps_per_second": 13.292,
"step": 5000
},
{
"epoch": 1.09,
"learning_rate": 3.911106711542269e-05,
"loss": 0.8746,
"step": 5500
},
{
"epoch": 1.19,
"learning_rate": 3.812116412591566e-05,
"loss": 0.8641,
"step": 6000
},
{
"epoch": 1.19,
"eval_COMMENT": {
"f1": 0.732877167342058,
"number": 6830,
"precision": 0.6708014106773683,
"recall": 0.8076134699853587
},
"eval_NAME": {
"f1": 0.8192514567458539,
"number": 8829,
"precision": 0.8106220201796208,
"recall": 0.8280665987088005
},
"eval_QTY": {
"f1": 0.9835020102592541,
"number": 7140,
"precision": 0.9736480922316771,
"recall": 0.9935574229691877
},
"eval_RANGE_END": {
"f1": 0.7563025210084033,
"number": 94,
"precision": 0.625,
"recall": 0.9574468085106383
},
"eval_UNIT": {
"f1": 0.9528070915998311,
"number": 5724,
"precision": 0.9219081849371018,
"recall": 0.9858490566037735
},
"eval_loss": 0.8873679041862488,
"eval_overall_accuracy": 0.8376435688750286,
"eval_overall_f1": 0.8636256522470964,
"eval_overall_precision": 0.8331114214269477,
"eval_overall_recall": 0.8964601460670231,
"eval_runtime": 20.1501,
"eval_samples_per_second": 422.131,
"eval_steps_per_second": 13.201,
"step": 6000
},
{
"epoch": 1.29,
"learning_rate": 3.7131261136408633e-05,
"loss": 0.8724,
"step": 6500
},
{
"epoch": 1.39,
"learning_rate": 3.6141358146901606e-05,
"loss": 0.8716,
"step": 7000
},
{
"epoch": 1.39,
"eval_COMMENT": {
"f1": 0.7425133689839571,
"number": 6830,
"precision": 0.6831488314883148,
"recall": 0.8131771595900439
},
"eval_NAME": {
"f1": 0.8184515478922914,
"number": 8829,
"precision": 0.8091653752490591,
"recall": 0.8279533355985955
},
"eval_QTY": {
"f1": 0.9851161496731118,
"number": 7140,
"precision": 0.9784470848300636,
"recall": 0.9918767507002801
},
"eval_RANGE_END": {
"f1": 0.7822222222222222,
"number": 94,
"precision": 0.6717557251908397,
"recall": 0.9361702127659575
},
"eval_UNIT": {
"f1": 0.9531475048716428,
"number": 5724,
"precision": 0.9253166639249877,
"recall": 0.9827044025157232
},
"eval_loss": 0.8783700466156006,
"eval_overall_accuracy": 0.8419981744884765,
"eval_overall_f1": 0.8664336726941195,
"eval_overall_precision": 0.8382007056056449,
"eval_overall_recall": 0.8966348673865185,
"eval_runtime": 19.9222,
"eval_samples_per_second": 426.962,
"eval_steps_per_second": 13.352,
"step": 7000
},
{
"epoch": 1.48,
"learning_rate": 3.515145515739457e-05,
"loss": 0.8674,
"step": 7500
},
{
"epoch": 1.58,
"learning_rate": 3.416155216788755e-05,
"loss": 0.8613,
"step": 8000
},
{
"epoch": 1.58,
"eval_COMMENT": {
"f1": 0.7415834380580726,
"number": 6830,
"precision": 0.6763180118228979,
"recall": 0.8207906295754026
},
"eval_NAME": {
"f1": 0.8138368983957219,
"number": 8829,
"precision": 0.8007234462347912,
"recall": 0.8273870200475705
},
"eval_QTY": {
"f1": 0.9851784844478463,
"number": 7140,
"precision": 0.9789793942746509,
"recall": 0.9914565826330533
},
"eval_RANGE_END": {
"f1": 0.7500000000000001,
"number": 94,
"precision": 0.6164383561643836,
"recall": 0.9574468085106383
},
"eval_UNIT": {
"f1": 0.9527118644067797,
"number": 5724,
"precision": 0.9251152073732719,
"recall": 0.9820055904961565
},
"eval_loss": 0.8823091983795166,
"eval_overall_accuracy": 0.8402487259450825,
"eval_overall_f1": 0.8641605863958844,
"eval_overall_precision": 0.832690750040499,
"eval_overall_recall": 0.8981025264702799,
"eval_runtime": 20.0298,
"eval_samples_per_second": 424.668,
"eval_steps_per_second": 13.28,
"step": 8000
},
{
"epoch": 1.68,
"learning_rate": 3.317164917838052e-05,
"loss": 0.8674,
"step": 8500
},
{
"epoch": 1.78,
"learning_rate": 3.218174618887349e-05,
"loss": 0.8744,
"step": 9000
},
{
"epoch": 1.78,
"eval_COMMENT": {
"f1": 0.7462050599201066,
"number": 6830,
"precision": 0.6842490842490843,
"recall": 0.820497803806735
},
"eval_NAME": {
"f1": 0.8154463390170511,
"number": 8829,
"precision": 0.8025666337611056,
"recall": 0.8287461773700305
},
"eval_QTY": {
"f1": 0.9861886160714286,
"number": 7140,
"precision": 0.9823513062812673,
"recall": 0.9900560224089636
},
"eval_RANGE_END": {
"f1": 0.7964601769911505,
"number": 94,
"precision": 0.6818181818181818,
"recall": 0.9574468085106383
},
"eval_UNIT": {
"f1": 0.954813690777272,
"number": 5724,
"precision": 0.922613229064842,
"recall": 0.9893431167016072
},
"eval_loss": 0.878822922706604,
"eval_overall_accuracy": 0.8422834106640298,
"eval_overall_f1": 0.8669136218218555,
"eval_overall_precision": 0.8365450232346537,
"eval_overall_recall": 0.8995701855540413,
"eval_runtime": 20.5948,
"eval_samples_per_second": 413.018,
"eval_steps_per_second": 12.916,
"step": 9000
},
{
"epoch": 1.88,
"learning_rate": 3.119184319936647e-05,
"loss": 0.8611,
"step": 9500
},
{
"epoch": 1.98,
"learning_rate": 3.0201940209859435e-05,
"loss": 0.8644,
"step": 10000
},
{
"epoch": 1.98,
"eval_COMMENT": {
"f1": 0.7429378531073446,
"number": 6830,
"precision": 0.6871112216969395,
"recall": 0.8086383601756955
},
"eval_NAME": {
"f1": 0.8145920356446671,
"number": 8829,
"precision": 0.8013368397983782,
"recall": 0.8282931249292106
},
"eval_QTY": {
"f1": 0.9860413176996092,
"number": 7140,
"precision": 0.9827490261547023,
"recall": 0.9893557422969188
},
"eval_RANGE_END": {
"f1": 0.7945205479452053,
"number": 94,
"precision": 0.696,
"recall": 0.925531914893617
},
"eval_UNIT": {
"f1": 0.9538357094365241,
"number": 5724,
"precision": 0.9273927392739274,
"recall": 0.9818308874912649
},
"eval_loss": 0.8753200769424438,
"eval_overall_accuracy": 0.8415988438427018,
"eval_overall_f1": 0.8657740812117525,
"eval_overall_precision": 0.8385565052231719,
"eval_overall_recall": 0.8948177656637663,
"eval_runtime": 20.5359,
"eval_samples_per_second": 414.201,
"eval_steps_per_second": 12.953,
"step": 10000
},
{
"epoch": 2.08,
"learning_rate": 2.9212037220352405e-05,
"loss": 0.8524,
"step": 10500
},
{
"epoch": 2.18,
"learning_rate": 2.8222134230845377e-05,
"loss": 0.8374,
"step": 11000
},
{
"epoch": 2.18,
"eval_COMMENT": {
"f1": 0.7445324666531248,
"number": 6830,
"precision": 0.6925305454087417,
"recall": 0.8049780380673499
},
"eval_NAME": {
"f1": 0.8132063916263014,
"number": 8829,
"precision": 0.7997152869031976,
"recall": 0.8271604938271605
},
"eval_QTY": {
"f1": 0.9866648048593173,
"number": 7140,
"precision": 0.9837115411388,
"recall": 0.9896358543417367
},
"eval_RANGE_END": {
"f1": 0.7665198237885463,
"number": 94,
"precision": 0.6541353383458647,
"recall": 0.925531914893617
},
"eval_UNIT": {
"f1": 0.9542107490478207,
"number": 5724,
"precision": 0.9254637990477754,
"recall": 0.9848008385744235
},
"eval_loss": 0.8822603821754456,
"eval_overall_accuracy": 0.8415227808625542,
"eval_overall_f1": 0.8660969625179796,
"eval_overall_precision": 0.839654832994291,
"eval_overall_recall": 0.894258657441381,
"eval_runtime": 19.8308,
"eval_samples_per_second": 428.929,
"eval_steps_per_second": 13.413,
"step": 11000
},
{
"epoch": 2.28,
"learning_rate": 2.7232231241338353e-05,
"loss": 0.8479,
"step": 11500
},
{
"epoch": 2.38,
"learning_rate": 2.6242328251831323e-05,
"loss": 0.8363,
"step": 12000
},
{
"epoch": 2.38,
"eval_COMMENT": {
"f1": 0.7471627157343361,
"number": 6830,
"precision": 0.6901128892196998,
"recall": 0.8144948755490483
},
"eval_NAME": {
"f1": 0.8170022371364652,
"number": 8829,
"precision": 0.8069826538504032,
"recall": 0.8272737569373655
},
"eval_QTY": {
"f1": 0.98654208214211,
"number": 7140,
"precision": 0.982363560616581,
"recall": 0.9907563025210084
},
"eval_RANGE_END": {
"f1": 0.7818181818181817,
"number": 94,
"precision": 0.6825396825396826,
"recall": 0.9148936170212766
},
"eval_UNIT": {
"f1": 0.9556608563208665,
"number": 5724,
"precision": 0.926649163111257,
"recall": 0.9865478686233403
},
"eval_loss": 0.886879026889801,
"eval_overall_accuracy": 0.8438807332471286,
"eval_overall_f1": 0.8680980557903636,
"eval_overall_precision": 0.8408607080863328,
"eval_overall_recall": 0.8971590313450047,
"eval_runtime": 20.4382,
"eval_samples_per_second": 416.181,
"eval_steps_per_second": 13.015,
"step": 12000
},
{
"epoch": 2.47,
"learning_rate": 2.5252425262324292e-05,
"loss": 0.8447,
"step": 12500
},
{
"epoch": 2.57,
"learning_rate": 2.4262522272817265e-05,
"loss": 0.8336,
"step": 13000
},
{
"epoch": 2.57,
"eval_COMMENT": {
"f1": 0.7385342789598108,
"number": 6830,
"precision": 0.6855172413793104,
"recall": 0.8004392386530015
},
"eval_NAME": {
"f1": 0.8182683158896289,
"number": 8829,
"precision": 0.8088072582429741,
"recall": 0.8279533355985955
},
"eval_QTY": {
"f1": 0.985723239779929,
"number": 7140,
"precision": 0.9803296855520155,
"recall": 0.9911764705882353
},
"eval_RANGE_END": {
"f1": 0.7733333333333333,
"number": 94,
"precision": 0.6641221374045801,
"recall": 0.925531914893617
},
"eval_UNIT": {
"f1": 0.954733903037482,
"number": 5724,
"precision": 0.9256767842493847,
"recall": 0.9856743535988819
},
"eval_loss": 0.8826401233673096,
"eval_overall_accuracy": 0.841560812352628,
"eval_overall_f1": 0.8661193398222599,
"eval_overall_precision": 0.8399435287937488,
"eval_overall_recall": 0.8939791033301884,
"eval_runtime": 21.5415,
"eval_samples_per_second": 394.867,
"eval_steps_per_second": 12.348,
"step": 13000
},
{
"epoch": 2.67,
"learning_rate": 2.3272619283310237e-05,
"loss": 0.8276,
"step": 13500
},
{
"epoch": 2.77,
"learning_rate": 2.2282716293803206e-05,
"loss": 0.8369,
"step": 14000
},
{
"epoch": 2.77,
"eval_COMMENT": {
"f1": 0.740984055293232,
"number": 6830,
"precision": 0.6956186560452268,
"recall": 0.792679355783309
},
"eval_NAME": {
"f1": 0.8157497905612958,
"number": 8829,
"precision": 0.8046496253856324,
"recall": 0.8271604938271605
},
"eval_QTY": {
"f1": 0.9862861120779673,
"number": 7140,
"precision": 0.980484429065744,
"recall": 0.9921568627450981
},
"eval_RANGE_END": {
"f1": 0.7787610619469028,
"number": 94,
"precision": 0.6666666666666666,
"recall": 0.9361702127659575
},
"eval_UNIT": {
"f1": 0.9553601623266824,
"number": 5724,
"precision": 0.9256225425950196,
"recall": 0.9870719776380154
},
"eval_loss": 0.8762015700340271,
"eval_overall_accuracy": 0.8427968357800258,
"eval_overall_f1": 0.8666542240019003,
"eval_overall_precision": 0.8423153034300792,
"eval_overall_recall": 0.8924415557186288,
"eval_runtime": 21.4465,
"eval_samples_per_second": 396.615,
"eval_steps_per_second": 12.403,
"step": 14000
},
{
"epoch": 2.87,
"learning_rate": 2.129281330429618e-05,
"loss": 0.8323,
"step": 14500
},
{
"epoch": 2.97,
"learning_rate": 2.0302910314789152e-05,
"loss": 0.8402,
"step": 15000
},
{
"epoch": 2.97,
"eval_COMMENT": {
"f1": 0.7500846367391156,
"number": 6830,
"precision": 0.6976949237939287,
"recall": 0.8109809663250366
},
"eval_NAME": {
"f1": 0.8182375564349813,
"number": 8829,
"precision": 0.805531167690957,
"recall": 0.8313512289047458
},
"eval_QTY": {
"f1": 0.9870909217779638,
"number": 7140,
"precision": 0.9835905993603115,
"recall": 0.9906162464985995
},
"eval_RANGE_END": {
"f1": 0.8173076923076923,
"number": 94,
"precision": 0.7456140350877193,
"recall": 0.9042553191489362
},
"eval_UNIT": {
"f1": 0.9550875412331896,
"number": 5724,
"precision": 0.9257255287752091,
"recall": 0.9863731656184487
},
"eval_loss": 0.8754065036773682,
"eval_overall_accuracy": 0.8451738039096371,
"eval_overall_f1": 0.8695490249187432,
"eval_overall_precision": 0.8433098013462486,
"eval_overall_recall": 0.8974735297200964,
"eval_runtime": 20.2196,
"eval_samples_per_second": 420.68,
"eval_steps_per_second": 13.156,
"step": 15000
},
{
"epoch": 3.07,
"learning_rate": 1.9313007325282124e-05,
"loss": 0.8134,
"step": 15500
},
{
"epoch": 3.17,
"learning_rate": 1.8323104335775097e-05,
"loss": 0.8036,
"step": 16000
},
{
"epoch": 3.17,
"eval_COMMENT": {
"f1": 0.7475294436171653,
"number": 6830,
"precision": 0.6951158106747231,
"recall": 0.8084919472913616
},
"eval_NAME": {
"f1": 0.8165045286816504,
"number": 8829,
"precision": 0.8062272275587943,
"recall": 0.8270472307169555
},
"eval_QTY": {
"f1": 0.9866127457816204,
"number": 7140,
"precision": 0.9823660094418217,
"recall": 0.9908963585434174
},
"eval_RANGE_END": {
"f1": 0.8038277511961722,
"number": 94,
"precision": 0.7304347826086957,
"recall": 0.8936170212765957
},
"eval_UNIT": {
"f1": 0.9559993230665087,
"number": 5724,
"precision": 0.9269773547751887,
"recall": 0.9868972746331237
},
"eval_loss": 0.884689450263977,
"eval_overall_accuracy": 0.8448315204989731,
"eval_overall_f1": 0.8684544884717681,
"eval_overall_precision": 0.8428252005787189,
"eval_overall_recall": 0.8956913722612433,
"eval_runtime": 22.2412,
"eval_samples_per_second": 382.444,
"eval_steps_per_second": 11.96,
"step": 16000
},
{
"epoch": 3.27,
"learning_rate": 1.7333201346268066e-05,
"loss": 0.8139,
"step": 16500
},
{
"epoch": 3.37,
"learning_rate": 1.634329835676104e-05,
"loss": 0.8037,
"step": 17000
},
{
"epoch": 3.37,
"eval_COMMENT": {
"f1": 0.7455275151350249,
"number": 6830,
"precision": 0.6962266548087918,
"recall": 0.8023426061493412
},
"eval_NAME": {
"f1": 0.8132627472833659,
"number": 8829,
"precision": 0.8004607283896445,
"recall": 0.8264809151659305
},
"eval_QTY": {
"f1": 0.9867019424911231,
"number": 7140,
"precision": 0.9810328118510314,
"recall": 0.992436974789916
},
"eval_RANGE_END": {
"f1": 0.8095238095238094,
"number": 94,
"precision": 0.7327586206896551,
"recall": 0.9042553191489362
},
"eval_UNIT": {
"f1": 0.9545839688188443,
"number": 5724,
"precision": 0.92678512668641,
"recall": 0.9841020265548568
},
"eval_loss": 0.8833730816841125,
"eval_overall_accuracy": 0.8431391191906898,
"eval_overall_f1": 0.8668440046763016,
"eval_overall_precision": 0.8413695566372845,
"eval_overall_recall": 0.8939092148023902,
"eval_runtime": 19.9144,
"eval_samples_per_second": 427.127,
"eval_steps_per_second": 13.357,
"step": 17000
},
{
"epoch": 3.46,
"learning_rate": 1.535339536725401e-05,
"loss": 0.8162,
"step": 17500
},
{
"epoch": 3.56,
"learning_rate": 1.4363492377746981e-05,
"loss": 0.8108,
"step": 18000
},
{
"epoch": 3.56,
"eval_COMMENT": {
"f1": 0.7473805178124789,
"number": 6830,
"precision": 0.694210724601281,
"recall": 0.8093704245973645
},
"eval_NAME": {
"f1": 0.8129492394272023,
"number": 8829,
"precision": 0.800065803904365,
"recall": 0.8262543889455204
},
"eval_QTY": {
"f1": 0.9860154456272177,
"number": 7140,
"precision": 0.9796764827872252,
"recall": 0.992436974789916
},
"eval_RANGE_END": {
"f1": 0.8036529680365296,
"number": 94,
"precision": 0.704,
"recall": 0.9361702127659575
},
"eval_UNIT": {
"f1": 0.9554840893703453,
"number": 5724,
"precision": 0.9266250820748523,
"recall": 0.9861984626135569
},
"eval_loss": 0.8875806927680969,
"eval_overall_accuracy": 0.8441279379326082,
"eval_overall_f1": 0.8670453776966254,
"eval_overall_precision": 0.8398676754773836,
"eval_overall_recall": 0.8960408149002341,
"eval_runtime": 20.6634,
"eval_samples_per_second": 411.645,
"eval_steps_per_second": 12.873,
"step": 18000
},
{
"epoch": 3.66,
"learning_rate": 1.3373589388239954e-05,
"loss": 0.8051,
"step": 18500
},
{
"epoch": 3.76,
"learning_rate": 1.2383686398732925e-05,
"loss": 0.806,
"step": 19000
},
{
"epoch": 3.76,
"eval_COMMENT": {
"f1": 0.7503922504945766,
"number": 6830,
"precision": 0.7025162856048026,
"recall": 0.8052708638360175
},
"eval_NAME": {
"f1": 0.819804195804196,
"number": 8829,
"precision": 0.8099712580145921,
"recall": 0.8298788084720806
},
"eval_QTY": {
"f1": 0.9867669591865162,
"number": 7140,
"precision": 0.9814353006372957,
"recall": 0.9921568627450981
},
"eval_RANGE_END": {
"f1": 0.7902439024390245,
"number": 94,
"precision": 0.7297297297297297,
"recall": 0.8617021276595744
},
"eval_UNIT": {
"f1": 0.9545956038360349,
"number": 5724,
"precision": 0.9282059745832646,
"recall": 0.9825296995108316
},
"eval_loss": 0.8832929134368896,
"eval_overall_accuracy": 0.8454210085951167,
"eval_overall_f1": 0.870108695652174,
"eval_overall_precision": 0.8464461553712455,
"eval_overall_recall": 0.895132264038858,
"eval_runtime": 19.9733,
"eval_samples_per_second": 425.868,
"eval_steps_per_second": 13.318,
"step": 19000
},
{
"epoch": 3.86,
"learning_rate": 1.1393783409225896e-05,
"loss": 0.8094,
"step": 19500
},
{
"epoch": 3.96,
"learning_rate": 1.0403880419718868e-05,
"loss": 0.8118,
"step": 20000
},
{
"epoch": 3.96,
"eval_COMMENT": {
"f1": 0.7498983326555511,
"number": 6830,
"precision": 0.6981322564361434,
"recall": 0.8099560761346999
},
"eval_NAME": {
"f1": 0.8188632803759232,
"number": 8829,
"precision": 0.8089974577207915,
"recall": 0.8289727035904406
},
"eval_QTY": {
"f1": 0.9865495853369572,
"number": 7140,
"precision": 0.9818282702177833,
"recall": 0.9913165266106443
},
"eval_RANGE_END": {
"f1": 0.7767857142857142,
"number": 94,
"precision": 0.6692307692307692,
"recall": 0.925531914893617
},
"eval_UNIT": {
"f1": 0.9557777024737377,
"number": 5724,
"precision": 0.9277960526315789,
"recall": 0.9854996505939903
},
"eval_loss": 0.880511462688446,
"eval_overall_accuracy": 0.8433482923860957,
"eval_overall_f1": 0.8696256376362127,
"eval_overall_precision": 0.8442579795985522,
"eval_overall_recall": 0.8965649788587203,
"eval_runtime": 19.9194,
"eval_samples_per_second": 427.021,
"eval_steps_per_second": 13.354,
"step": 20000
},
{
"epoch": 4.06,
"learning_rate": 9.41397743021184e-06,
"loss": 0.7833,
"step": 20500
},
{
"epoch": 4.16,
"learning_rate": 8.424074440704812e-06,
"loss": 0.7792,
"step": 21000
},
{
"epoch": 4.16,
"eval_COMMENT": {
"f1": 0.7491693225740829,
"number": 6830,
"precision": 0.6977390425666288,
"recall": 0.8087847730600293
},
"eval_NAME": {
"f1": 0.8171393410527494,
"number": 8829,
"precision": 0.8072502210433244,
"recall": 0.8272737569373655
},
"eval_QTY": {
"f1": 0.9866239375783753,
"number": 7140,
"precision": 0.981563626282229,
"recall": 0.9917366946778712
},
"eval_RANGE_END": {
"f1": 0.7945205479452053,
"number": 94,
"precision": 0.696,
"recall": 0.925531914893617
},
"eval_UNIT": {
"f1": 0.955064737242955,
"number": 5724,
"precision": 0.9261447562776958,
"recall": 0.9858490566037735
},
"eval_loss": 0.8954847455024719,
"eval_overall_accuracy": 0.843823686012018,
"eval_overall_f1": 0.8689124614498255,
"eval_overall_precision": 0.843471395203474,
"eval_overall_recall": 0.8959359821085369,
"eval_runtime": 22.6334,
"eval_samples_per_second": 375.816,
"eval_steps_per_second": 11.753,
"step": 21000
},
{
"epoch": 4.26,
"learning_rate": 7.434171451197783e-06,
"loss": 0.7828,
"step": 21500
},
{
"epoch": 4.36,
"learning_rate": 6.444268461690754e-06,
"loss": 0.7844,
"step": 22000
},
{
"epoch": 4.36,
"eval_COMMENT": {
"f1": 0.7466903234611709,
"number": 6830,
"precision": 0.6992586912065439,
"recall": 0.8010248901903367
},
"eval_NAME": {
"f1": 0.8147858618571668,
"number": 8829,
"precision": 0.8035242290748899,
"recall": 0.8263676520557255
},
"eval_QTY": {
"f1": 0.9864234491401519,
"number": 7140,
"precision": 0.9807559185933823,
"recall": 0.9921568627450981
},
"eval_RANGE_END": {
"f1": 0.7999999999999999,
"number": 94,
"precision": 0.7107438016528925,
"recall": 0.9148936170212766
},
"eval_UNIT": {
"f1": 0.9547645218567684,
"number": 5724,
"precision": 0.9251187940357202,
"recall": 0.9863731656184487
},
"eval_loss": 0.8964929580688477,
"eval_overall_accuracy": 0.8420362059785502,
"eval_overall_f1": 0.8676909510242844,
"eval_overall_precision": 0.8429046818885704,
"eval_overall_recall": 0.8939791033301884,
"eval_runtime": 21.3934,
"eval_samples_per_second": 397.598,
"eval_steps_per_second": 12.434,
"step": 22000
},
{
"epoch": 4.45,
"learning_rate": 5.4543654721837265e-06,
"loss": 0.7805,
"step": 22500
},
{
"epoch": 4.55,
"learning_rate": 4.4644624826766974e-06,
"loss": 0.7783,
"step": 23000
},
{
"epoch": 4.55,
"eval_COMMENT": {
"f1": 0.7446431008598335,
"number": 6830,
"precision": 0.6973415132924335,
"recall": 0.7988286969253294
},
"eval_NAME": {
"f1": 0.8151997764738753,
"number": 8829,
"precision": 0.8045444517979263,
"recall": 0.8261411258353154
},
"eval_QTY": {
"f1": 0.9871053181849865,
"number": 7140,
"precision": 0.9825169973636743,
"recall": 0.9917366946778712
},
"eval_RANGE_END": {
"f1": 0.7853881278538812,
"number": 94,
"precision": 0.688,
"recall": 0.9148936170212766
},
"eval_UNIT": {
"f1": 0.9550219817382481,
"number": 5724,
"precision": 0.9252948885976409,
"recall": 0.986722571628232
},
"eval_loss": 0.8985583186149597,
"eval_overall_accuracy": 0.8421503004487716,
"eval_overall_f1": 0.8674482126800468,
"eval_overall_precision": 0.8430060014509002,
"eval_overall_recall": 0.8933501065800049,
"eval_runtime": 22.3207,
"eval_samples_per_second": 381.081,
"eval_steps_per_second": 11.917,
"step": 23000
},
{
"epoch": 4.65,
"learning_rate": 3.4745594931696697e-06,
"loss": 0.7777,
"step": 23500
},
{
"epoch": 4.75,
"learning_rate": 2.484656503662641e-06,
"loss": 0.784,
"step": 24000
},
{
"epoch": 4.75,
"eval_COMMENT": {
"f1": 0.7473469387755102,
"number": 6830,
"precision": 0.6979669631512071,
"recall": 0.8042459736456808
},
"eval_NAME": {
"f1": 0.8153889043769166,
"number": 8829,
"precision": 0.8029870415110916,
"recall": 0.8281798618190056
},
"eval_QTY": {
"f1": 0.9870401337792643,
"number": 7140,
"precision": 0.9821131447587355,
"recall": 0.9920168067226891
},
"eval_RANGE_END": {
"f1": 0.7889908256880733,
"number": 94,
"precision": 0.6935483870967742,
"recall": 0.9148936170212766
},
"eval_UNIT": {
"f1": 0.9550419100838201,
"number": 5724,
"precision": 0.9265648102513554,
"recall": 0.9853249475890985
},
"eval_loss": 0.8965936303138733,
"eval_overall_accuracy": 0.8442800638929033,
"eval_overall_f1": 0.868035786905246,
"eval_overall_precision": 0.8425935063653409,
"eval_overall_recall": 0.8950623755110598,
"eval_runtime": 20.0824,
"eval_samples_per_second": 423.555,
"eval_steps_per_second": 13.245,
"step": 24000
},
{
"epoch": 4.85,
"learning_rate": 1.4947535141556129e-06,
"loss": 0.7761,
"step": 24500
},
{
"epoch": 4.95,
"learning_rate": 5.048505246485845e-07,
"loss": 0.776,
"step": 25000
},
{
"epoch": 4.95,
"eval_COMMENT": {
"f1": 0.7472796517954298,
"number": 6830,
"precision": 0.6977393954787909,
"recall": 0.8043923865300147
},
"eval_NAME": {
"f1": 0.8170697648438808,
"number": 8829,
"precision": 0.8060392329733304,
"recall": 0.8284063880394156
},
"eval_QTY": {
"f1": 0.9868971285196543,
"number": 7140,
"precision": 0.9822419533851277,
"recall": 0.9915966386554622
},
"eval_RANGE_END": {
"f1": 0.7926267281105991,
"number": 94,
"precision": 0.6991869918699187,
"recall": 0.9148936170212766
},
"eval_UNIT": {
"f1": 0.9552415601996784,
"number": 5724,
"precision": 0.92616899097621,
"recall": 0.9861984626135569
},
"eval_loss": 0.8963901400566101,
"eval_overall_accuracy": 0.8442040009127557,
"eval_overall_f1": 0.8685731721788069,
"eval_overall_precision": 0.8434516362678607,
"eval_overall_recall": 0.8952370968305553,
"eval_runtime": 19.5692,
"eval_samples_per_second": 434.663,
"eval_steps_per_second": 13.593,
"step": 25000
},
{
"epoch": 5.0,
"step": 25255,
"total_flos": 1.2443566938551148e+16,
"train_loss": 0.8428581520251249,
"train_runtime": 5937.5702,
"train_samples_per_second": 136.085,
"train_steps_per_second": 4.253
}
],
"logging_steps": 500,
"max_steps": 25255,
"num_train_epochs": 5,
"save_steps": 1000,
"total_flos": 1.2443566938551148e+16,
"trial_name": null,
"trial_params": null
}