longformer-spans / checkpoint-729 /trainer_state.json
Theoreticallyhugo's picture
Training in progress, epoch 9, checkpoint
70cbb60 verified
raw
history blame
10.7 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 9.0,
"eval_steps": 500,
"global_step": 729,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"eval_B": {
"f1-score": 0.8331822302810515,
"precision": 0.7901977644024075,
"recall": 0.8811121764141898,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.9471294753175105,
"precision": 0.9443616777446711,
"recall": 0.9499135446685879,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.8992868897421833,
"precision": 0.9103731674811195,
"recall": 0.8884673748103187,
"support": 9226.0
},
"eval_accuracy": 0.9267895289474637,
"eval_loss": 0.20478281378746033,
"eval_macro avg": {
"f1-score": 0.8931995317802484,
"precision": 0.8816442032093993,
"recall": 0.9064976986310321,
"support": 27619.0
},
"eval_runtime": 4.6128,
"eval_samples_per_second": 17.343,
"eval_steps_per_second": 2.168,
"eval_weighted avg": {
"f1-score": 0.9268447919078652,
"precision": 0.9271861479533134,
"recall": 0.9267895289474637,
"support": 27619.0
},
"step": 81
},
{
"epoch": 2.0,
"eval_B": {
"f1-score": 0.8682547399124939,
"precision": 0.8806706114398422,
"recall": 0.8561840843720039,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.9590983771942466,
"precision": 0.9592919333448654,
"recall": 0.9589048991354466,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.9226525313717006,
"precision": 0.9208594256100194,
"recall": 0.9244526338608281,
"support": 9226.0
},
"eval_accuracy": 0.9435171439950758,
"eval_loss": 0.16574698686599731,
"eval_macro avg": {
"f1-score": 0.9166685494928136,
"precision": 0.9202739901315756,
"recall": 0.9131805391227594,
"support": 27619.0
},
"eval_runtime": 4.6238,
"eval_samples_per_second": 17.302,
"eval_steps_per_second": 2.163,
"eval_weighted avg": {
"f1-score": 0.9434932036816766,
"precision": 0.9434846863370583,
"recall": 0.9435171439950758,
"support": 27619.0
},
"step": 162
},
{
"epoch": 3.0,
"eval_B": {
"f1-score": 0.8791725434884815,
"precision": 0.8625461254612546,
"recall": 0.8964525407478428,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.9588372159825352,
"precision": 0.9557349825345015,
"recall": 0.9619596541786744,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.9204284621270084,
"precision": 0.9282407407407407,
"recall": 0.9127465857359636,
"support": 9226.0
},
"eval_accuracy": 0.9430464535283681,
"eval_loss": 0.15394224226474762,
"eval_macro avg": {
"f1-score": 0.9194794071993417,
"precision": 0.9155072829121655,
"recall": 0.9237195935541602,
"support": 27619.0
},
"eval_runtime": 4.6233,
"eval_samples_per_second": 17.303,
"eval_steps_per_second": 2.163,
"eval_weighted avg": {
"f1-score": 0.9429985029052194,
"precision": 0.9430314866542512,
"recall": 0.9430464535283681,
"support": 27619.0
},
"step": 243
},
{
"epoch": 4.0,
"eval_B": {
"f1-score": 0.8800746616892207,
"precision": 0.8572727272727273,
"recall": 0.9041227229146692,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.9537844036697248,
"precision": 0.9488876212207644,
"recall": 0.9587319884726225,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.9100192149327477,
"precision": 0.9220157970853265,
"recall": 0.8983308042488619,
"support": 9226.0
},
"eval_accuracy": 0.9364929939534379,
"eval_loss": 0.19801080226898193,
"eval_macro avg": {
"f1-score": 0.9146260934305644,
"precision": 0.9093920485262728,
"recall": 0.9203951718787179,
"support": 27619.0
},
"eval_runtime": 4.6132,
"eval_samples_per_second": 17.342,
"eval_steps_per_second": 2.168,
"eval_weighted avg": {
"f1-score": 0.9363812792925563,
"precision": 0.9364514800186444,
"recall": 0.9364929939534379,
"support": 27619.0
},
"step": 324
},
{
"epoch": 5.0,
"eval_B": {
"f1-score": 0.883877159309021,
"precision": 0.8847262247838616,
"recall": 0.8830297219558965,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.9591500482927107,
"precision": 0.9456643513331839,
"recall": 0.9730259365994236,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.9191176470588236,
"precision": 0.9454503781801513,
"recall": 0.8942120095382614,
"support": 9226.0
},
"eval_accuracy": 0.9432999022412107,
"eval_loss": 0.20074255764484406,
"eval_macro avg": {
"f1-score": 0.9207149515535183,
"precision": 0.9252803180990656,
"recall": 0.9167558893645271,
"support": 27619.0
},
"eval_runtime": 4.6092,
"eval_samples_per_second": 17.356,
"eval_steps_per_second": 2.17,
"eval_weighted avg": {
"f1-score": 0.9429348139614955,
"precision": 0.9432916158141273,
"recall": 0.9432999022412107,
"support": 27619.0
},
"step": 405
},
{
"epoch": 6.0,
"eval_B": {
"f1-score": 0.8897338403041826,
"precision": 0.882186616399623,
"recall": 0.8974113135186961,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.9624160683661308,
"precision": 0.9707969272268809,
"recall": 0.954178674351585,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.9297955261331482,
"precision": 0.9161493950552341,
"recall": 0.9438543247344461,
"support": 9226.0
},
"eval_accuracy": 0.9485861182519281,
"eval_loss": 0.16565027832984924,
"eval_macro avg": {
"f1-score": 0.9273151449344872,
"precision": 0.9230443128939126,
"recall": 0.9318147708682424,
"support": 27619.0
},
"eval_runtime": 4.6219,
"eval_samples_per_second": 17.309,
"eval_steps_per_second": 2.164,
"eval_weighted avg": {
"f1-score": 0.9487745648174827,
"precision": 0.9491959030765336,
"recall": 0.9485861182519281,
"support": 27619.0
},
"step": 486
},
{
"epoch": 6.17,
"grad_norm": 4.767655849456787,
"learning_rate": 1.7530864197530865e-05,
"loss": 0.1585,
"step": 500
},
{
"epoch": 7.0,
"eval_B": {
"f1-score": 0.8802588996763754,
"precision": 0.85,
"recall": 0.912751677852349,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.958423523458803,
"precision": 0.9354661691269275,
"recall": 0.9825360230547551,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.9139526911210146,
"precision": 0.9664088931851136,
"recall": 0.8668978972469109,
"support": 9226.0
},
"eval_accuracy": 0.9412723125384699,
"eval_loss": 0.2591131329536438,
"eval_macro avg": {
"f1-score": 0.9175450380853977,
"precision": 0.917291687437347,
"recall": 0.9207285327180049,
"support": 27619.0
},
"eval_runtime": 4.617,
"eval_samples_per_second": 17.327,
"eval_steps_per_second": 2.166,
"eval_weighted avg": {
"f1-score": 0.9406164485555297,
"precision": 0.9425749115781907,
"recall": 0.9412723125384699,
"support": 27619.0
},
"step": 567
},
{
"epoch": 8.0,
"eval_B": {
"f1-score": 0.8687910028116214,
"precision": 0.84967919340055,
"recall": 0.8887823585810163,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.9595010203788336,
"precision": 0.956997878562009,
"recall": 0.9620172910662824,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.9212035166275324,
"precision": 0.9282491471332673,
"recall": 0.9142640364188164,
"support": 9226.0
},
"eval_accuracy": 0.9432999022412107,
"eval_loss": 0.2693409323692322,
"eval_macro avg": {
"f1-score": 0.9164985132726624,
"precision": 0.9116420730319422,
"recall": 0.9216878953553717,
"support": 27619.0
},
"eval_runtime": 4.6153,
"eval_samples_per_second": 17.333,
"eval_steps_per_second": 2.167,
"eval_weighted avg": {
"f1-score": 0.9432823550422136,
"precision": 0.9433417293609165,
"recall": 0.9432999022412107,
"support": 27619.0
},
"step": 648
},
{
"epoch": 9.0,
"eval_B": {
"f1-score": 0.8645343367826904,
"precision": 0.8485687903970452,
"recall": 0.8811121764141898,
"support": 1043.0
},
"eval_I": {
"f1-score": 0.9567003348057282,
"precision": 0.934535260814599,
"recall": 0.9799423631123919,
"support": 17350.0
},
"eval_O": {
"f1-score": 0.9104673003585861,
"precision": 0.9586479683567062,
"recall": 0.8668978972469109,
"support": 9226.0
},
"eval_accuracy": 0.9384481697382236,
"eval_loss": 0.3087053894996643,
"eval_macro avg": {
"f1-score": 0.910567323982335,
"precision": 0.9139173398561168,
"recall": 0.9093174789244975,
"support": 27619.0
},
"eval_runtime": 4.6156,
"eval_samples_per_second": 17.333,
"eval_steps_per_second": 2.167,
"eval_weighted avg": {
"f1-score": 0.9377758584761231,
"precision": 0.9393435743356523,
"recall": 0.9384481697382236,
"support": 27619.0
},
"step": 729
}
],
"logging_steps": 500,
"max_steps": 4050,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"total_flos": 1294079874620400.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}