m0_flat_ner_ref_cmbert_io / trainer_state.json
nlpso's picture
Load model and tokenizer
e44045f
{
"best_metric": 0.9778851514588367,
"best_model_checkpoint": "/work/stual/res_ICDAR/method_0/tmp/21-flat-ner-ref-camembert_ner/checkpoint-1200",
"epoch": 3.1496062992125986,
"global_step": 1200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.26,
"eval_ACT": {
"f1": 0.9104618284637135,
"number": 519,
"precision": 0.8911439114391144,
"recall": 0.930635838150289
},
"eval_CARDINAL": {
"f1": 0.9919058130978661,
"number": 678,
"precision": 0.9897209985315712,
"recall": 0.9941002949852508
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.939820742637644,
"number": 761,
"precision": 0.916354556803995,
"recall": 0.9645203679369251
},
"eval_PER": {
"f1": 0.9195906432748537,
"number": 679,
"precision": 0.9129172714078374,
"recall": 0.9263622974963182
},
"eval_TITRE": {
"f1": 0.0,
"number": 40,
"precision": 0.0,
"recall": 0.0
},
"eval_accuracy": 0.9574230387288977,
"eval_f1": 0.933852140077821,
"eval_loss": 0.20205183327198029,
"eval_precision": 0.9288610394397346,
"eval_recall": 0.9388971684053651,
"eval_runtime": 2.1249,
"eval_samples_per_second": 318.138,
"eval_steps_per_second": 20.237,
"step": 100
},
{
"epoch": 0.52,
"eval_ACT": {
"f1": 0.9073900841908326,
"number": 519,
"precision": 0.8818181818181818,
"recall": 0.9344894026974951
},
"eval_CARDINAL": {
"f1": 0.9911764705882352,
"number": 678,
"precision": 0.9882697947214076,
"recall": 0.9941002949852508
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.943058221369162,
"number": 761,
"precision": 0.9189526184538653,
"recall": 0.9684625492772667
},
"eval_PER": {
"f1": 0.9545454545454546,
"number": 679,
"precision": 0.9503649635036496,
"recall": 0.9587628865979382
},
"eval_TITRE": {
"f1": 0.53125,
"number": 40,
"precision": 0.7083333333333334,
"recall": 0.425
},
"eval_accuracy": 0.961643495531281,
"eval_f1": 0.9449051041090842,
"eval_loss": 0.16001582145690918,
"eval_precision": 0.9347429821363471,
"eval_recall": 0.9552906110283159,
"eval_runtime": 2.1373,
"eval_samples_per_second": 316.284,
"eval_steps_per_second": 20.119,
"step": 200
},
{
"epoch": 0.79,
"eval_ACT": {
"f1": 0.9244019138755981,
"number": 519,
"precision": 0.9182509505703422,
"recall": 0.930635838150289
},
"eval_CARDINAL": {
"f1": 0.9926362297496317,
"number": 678,
"precision": 0.9911764705882353,
"recall": 0.9941002949852508
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.9435695538057742,
"number": 761,
"precision": 0.9423328964613368,
"recall": 0.9448094612352168
},
"eval_PER": {
"f1": 0.979381443298969,
"number": 679,
"precision": 0.979381443298969,
"recall": 0.979381443298969
},
"eval_TITRE": {
"f1": 0.8684210526315789,
"number": 40,
"precision": 0.9166666666666666,
"recall": 0.825
},
"eval_accuracy": 0.9690913604766633,
"eval_f1": 0.9590163934426229,
"eval_loss": 0.1148175373673439,
"eval_precision": 0.9590163934426229,
"eval_recall": 0.9590163934426229,
"eval_runtime": 2.1137,
"eval_samples_per_second": 319.821,
"eval_steps_per_second": 20.344,
"step": 300
},
{
"epoch": 1.05,
"eval_ACT": {
"f1": 0.9422894985808893,
"number": 519,
"precision": 0.9256505576208178,
"recall": 0.9595375722543352
},
"eval_CARDINAL": {
"f1": 0.9941089837997055,
"number": 678,
"precision": 0.9926470588235294,
"recall": 0.995575221238938
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.9652002626395274,
"number": 761,
"precision": 0.9645669291338582,
"recall": 0.9658344283837057
},
"eval_PER": {
"f1": 0.986784140969163,
"number": 679,
"precision": 0.9838945827232797,
"recall": 0.9896907216494846
},
"eval_TITRE": {
"f1": 0.8433734939759036,
"number": 40,
"precision": 0.813953488372093,
"recall": 0.875
},
"eval_accuracy": 0.977780536246276,
"eval_f1": 0.9703153988868275,
"eval_loss": 0.08161204308271408,
"eval_precision": 0.9663710273466372,
"eval_recall": 0.9742921013412816,
"eval_runtime": 2.1472,
"eval_samples_per_second": 314.828,
"eval_steps_per_second": 20.026,
"step": 400
},
{
"epoch": 1.31,
"learning_rate": 9e-05,
"loss": 0.2031,
"step": 500
},
{
"epoch": 1.31,
"eval_ACT": {
"f1": 0.9473684210526315,
"number": 519,
"precision": 0.94106463878327,
"recall": 0.953757225433526
},
"eval_CARDINAL": {
"f1": 0.9911634756995582,
"number": 678,
"precision": 0.9897058823529412,
"recall": 0.9926253687315634
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.9559014267185473,
"number": 761,
"precision": 0.9436619718309859,
"recall": 0.9684625492772667
},
"eval_PER": {
"f1": 0.9730517115804806,
"number": 679,
"precision": 0.962536023054755,
"recall": 0.9837997054491899
},
"eval_TITRE": {
"f1": 0.8536585365853658,
"number": 40,
"precision": 0.8333333333333334,
"recall": 0.875
},
"eval_accuracy": 0.9728152929493545,
"eval_f1": 0.9646754207508785,
"eval_loss": 0.10748306661844254,
"eval_precision": 0.9577671685640837,
"eval_recall": 0.9716840536512668,
"eval_runtime": 2.1738,
"eval_samples_per_second": 310.97,
"eval_steps_per_second": 19.781,
"step": 500
},
{
"epoch": 1.57,
"eval_ACT": {
"f1": 0.9438629876308278,
"number": 519,
"precision": 0.9323308270676691,
"recall": 0.9556840077071291
},
"eval_CARDINAL": {
"f1": 0.9919177075679647,
"number": 678,
"precision": 0.9882869692532943,
"recall": 0.995575221238938
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.9646596858638743,
"number": 761,
"precision": 0.9608865710560626,
"recall": 0.9684625492772667
},
"eval_PER": {
"f1": 0.9816311535635561,
"number": 679,
"precision": 0.9794721407624634,
"recall": 0.9837997054491899
},
"eval_TITRE": {
"f1": 0.8533333333333333,
"number": 40,
"precision": 0.9142857142857143,
"recall": 0.8
},
"eval_accuracy": 0.977780536246276,
"eval_f1": 0.9689764072078766,
"eval_loss": 0.08254563808441162,
"eval_precision": 0.9662838088180807,
"eval_recall": 0.9716840536512668,
"eval_runtime": 3.17,
"eval_samples_per_second": 213.252,
"eval_steps_per_second": 13.565,
"step": 600
},
{
"epoch": 1.84,
"eval_ACT": {
"f1": 0.958930276981853,
"number": 519,
"precision": 0.9507575757575758,
"recall": 0.9672447013487476
},
"eval_CARDINAL": {
"f1": 0.9919058130978661,
"number": 678,
"precision": 0.9897209985315712,
"recall": 0.9941002949852508
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.9653368214519294,
"number": 761,
"precision": 0.9609375,
"recall": 0.9697766097240473
},
"eval_PER": {
"f1": 0.9801616458486406,
"number": 679,
"precision": 0.9780058651026393,
"recall": 0.9823269513991163
},
"eval_TITRE": {
"f1": 0.7999999999999998,
"number": 40,
"precision": 0.7555555555555555,
"recall": 0.85
},
"eval_accuracy": 0.9781529294935452,
"eval_f1": 0.9706755753526355,
"eval_loss": 0.07011305540800095,
"eval_precision": 0.9670857988165681,
"eval_recall": 0.9742921013412816,
"eval_runtime": 2.1365,
"eval_samples_per_second": 316.402,
"eval_steps_per_second": 20.126,
"step": 700
},
{
"epoch": 2.1,
"eval_ACT": {
"f1": 0.9465648854961832,
"number": 519,
"precision": 0.9376181474480151,
"recall": 0.9556840077071291
},
"eval_CARDINAL": {
"f1": 0.9919058130978661,
"number": 678,
"precision": 0.9897209985315712,
"recall": 0.9941002949852508
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.9744931327665142,
"number": 761,
"precision": 0.9700520833333334,
"recall": 0.9789750328515112
},
"eval_PER": {
"f1": 0.9714285714285714,
"number": 679,
"precision": 0.9664723032069971,
"recall": 0.9764359351988218
},
"eval_TITRE": {
"f1": 0.8705882352941177,
"number": 40,
"precision": 0.8222222222222222,
"recall": 0.925
},
"eval_accuracy": 0.9767874875868917,
"eval_f1": 0.9697756350825143,
"eval_loss": 0.09748248755931854,
"eval_precision": 0.9653008490217793,
"eval_recall": 0.9742921013412816,
"eval_runtime": 2.1373,
"eval_samples_per_second": 316.283,
"eval_steps_per_second": 20.119,
"step": 800
},
{
"epoch": 2.36,
"eval_ACT": {
"f1": 0.9579349904397705,
"number": 519,
"precision": 0.9506641366223909,
"recall": 0.9653179190751445
},
"eval_CARDINAL": {
"f1": 0.9933774834437086,
"number": 678,
"precision": 0.9911894273127754,
"recall": 0.995575221238938
},
"eval_FT": {
"f1": 0.36363636363636365,
"number": 7,
"precision": 0.5,
"recall": 0.2857142857142857
},
"eval_LOC": {
"f1": 0.958957654723127,
"number": 761,
"precision": 0.9509043927648578,
"recall": 0.9671484888304862
},
"eval_PER": {
"f1": 0.9809663250366032,
"number": 679,
"precision": 0.975254730713246,
"recall": 0.9867452135493373
},
"eval_TITRE": {
"f1": 0.9135802469135802,
"number": 40,
"precision": 0.9024390243902439,
"recall": 0.925
},
"eval_accuracy": 0.9793942403177756,
"eval_f1": 0.9711004075583549,
"eval_loss": 0.07152657955884933,
"eval_precision": 0.9657332350773765,
"eval_recall": 0.9765275707898659,
"eval_runtime": 2.2181,
"eval_samples_per_second": 304.768,
"eval_steps_per_second": 19.386,
"step": 900
},
{
"epoch": 2.62,
"learning_rate": 8e-05,
"loss": 0.0707,
"step": 1000
},
{
"epoch": 2.62,
"eval_ACT": {
"f1": 0.9502868068833653,
"number": 519,
"precision": 0.9430740037950665,
"recall": 0.9576107899807321
},
"eval_CARDINAL": {
"f1": 0.9918938835666913,
"number": 678,
"precision": 0.9911634756995582,
"recall": 0.9926253687315634
},
"eval_FT": {
"f1": 0.0,
"number": 7,
"precision": 0.0,
"recall": 0.0
},
"eval_LOC": {
"f1": 0.9687499999999999,
"number": 761,
"precision": 0.96,
"recall": 0.9776609724047306
},
"eval_PER": {
"f1": 0.9889624724061811,
"number": 679,
"precision": 0.9882352941176471,
"recall": 0.9896907216494846
},
"eval_TITRE": {
"f1": 0.9,
"number": 40,
"precision": 0.9,
"recall": 0.9
},
"eval_accuracy": 0.9766633565044687,
"eval_f1": 0.9736353509097662,
"eval_loss": 0.08077961951494217,
"eval_precision": 0.9703923019985197,
"eval_recall": 0.9769001490312966,
"eval_runtime": 2.0842,
"eval_samples_per_second": 324.34,
"eval_steps_per_second": 20.631,
"step": 1000
},
{
"epoch": 2.89,
"eval_ACT": {
"f1": 0.9429657794676807,
"number": 519,
"precision": 0.9305816135084428,
"recall": 0.9556840077071291
},
"eval_CARDINAL": {
"f1": 0.9941089837997055,
"number": 678,
"precision": 0.9926470588235294,
"recall": 0.995575221238938
},
"eval_FT": {
"f1": 0.30769230769230765,
"number": 7,
"precision": 0.3333333333333333,
"recall": 0.2857142857142857
},
"eval_LOC": {
"f1": 0.9596354166666666,
"number": 761,
"precision": 0.9509677419354838,
"recall": 0.9684625492772667
},
"eval_PER": {
"f1": 0.9904341427520236,
"number": 679,
"precision": 0.9897058823529412,
"recall": 0.9911634756995582
},
"eval_TITRE": {
"f1": 0.9,
"number": 40,
"precision": 0.9,
"recall": 0.9
},
"eval_accuracy": 0.9740566037735849,
"eval_f1": 0.9703593923675434,
"eval_loss": 0.08018206059932709,
"eval_precision": 0.9649963154016212,
"eval_recall": 0.9757824143070045,
"eval_runtime": 2.2649,
"eval_samples_per_second": 298.463,
"eval_steps_per_second": 18.985,
"step": 1100
},
{
"epoch": 3.15,
"eval_ACT": {
"f1": 0.9636711281070746,
"number": 519,
"precision": 0.9563567362428842,
"recall": 0.9710982658959537
},
"eval_CARDINAL": {
"f1": 0.9941089837997055,
"number": 678,
"precision": 0.9926470588235294,
"recall": 0.995575221238938
},
"eval_FT": {
"f1": 0.4615384615384615,
"number": 7,
"precision": 0.5,
"recall": 0.42857142857142855
},
"eval_LOC": {
"f1": 0.9711286089238845,
"number": 761,
"precision": 0.9698558322411533,
"recall": 0.9724047306176085
},
"eval_PER": {
"f1": 0.9897058823529413,
"number": 679,
"precision": 0.9882525697503671,
"recall": 0.9911634756995582
},
"eval_TITRE": {
"f1": 0.9,
"number": 40,
"precision": 0.9,
"recall": 0.9
},
"eval_accuracy": 0.9822492552135055,
"eval_f1": 0.9778851514588367,
"eval_loss": 0.07905712723731995,
"eval_precision": 0.9755283648498332,
"eval_recall": 0.9802533532041728,
"eval_runtime": 2.1617,
"eval_samples_per_second": 312.718,
"eval_steps_per_second": 19.892,
"step": 1200
}
],
"max_steps": 5000,
"num_train_epochs": 14,
"total_flos": 486838277989008.0,
"trial_name": null,
"trial_params": null
}