Trisham97's picture
Add fine-tuned energy NER model
fb6c332 verified
{
"best_global_step": 1144,
"best_metric": 0.6738893476465732,
"best_model_checkpoint": "energy_intelligence_multitask_custom_ner/checkpoint-1144",
"epoch": 4.0,
"eval_steps": 500,
"global_step": 1144,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.06993006993006994,
"grad_norm": 14.66324520111084,
"learning_rate": 2.6573426573426574e-06,
"loss": 2.297128105163574,
"step": 20
},
{
"epoch": 0.13986013986013987,
"grad_norm": 8.70977783203125,
"learning_rate": 5.4545454545454545e-06,
"loss": 1.5782511711120606,
"step": 40
},
{
"epoch": 0.2097902097902098,
"grad_norm": 0.9631044268608093,
"learning_rate": 8.251748251748254e-06,
"loss": 0.7814332008361816,
"step": 60
},
{
"epoch": 0.27972027972027974,
"grad_norm": 0.513141393661499,
"learning_rate": 1.1048951048951048e-05,
"loss": 0.4949788570404053,
"step": 80
},
{
"epoch": 0.34965034965034963,
"grad_norm": 0.3357048034667969,
"learning_rate": 1.3846153846153847e-05,
"loss": 0.4088569164276123,
"step": 100
},
{
"epoch": 0.4195804195804196,
"grad_norm": 0.44831985235214233,
"learning_rate": 1.6643356643356645e-05,
"loss": 0.3331140518188477,
"step": 120
},
{
"epoch": 0.48951048951048953,
"grad_norm": 0.38962382078170776,
"learning_rate": 1.944055944055944e-05,
"loss": 0.28518569469451904,
"step": 140
},
{
"epoch": 0.5594405594405595,
"grad_norm": 0.550012469291687,
"learning_rate": 1.9751359751359752e-05,
"loss": 0.255126428604126,
"step": 160
},
{
"epoch": 0.6293706293706294,
"grad_norm": 0.4113660752773285,
"learning_rate": 1.944055944055944e-05,
"loss": 0.23595545291900635,
"step": 180
},
{
"epoch": 0.6993006993006993,
"grad_norm": 0.5528178811073303,
"learning_rate": 1.912975912975913e-05,
"loss": 0.23385438919067383,
"step": 200
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.5938523411750793,
"learning_rate": 1.8818958818958822e-05,
"loss": 0.217702054977417,
"step": 220
},
{
"epoch": 0.8391608391608392,
"grad_norm": 0.5627780556678772,
"learning_rate": 1.850815850815851e-05,
"loss": 0.20575783252716065,
"step": 240
},
{
"epoch": 0.9090909090909091,
"grad_norm": 0.6800552606582642,
"learning_rate": 1.81973581973582e-05,
"loss": 0.20057928562164307,
"step": 260
},
{
"epoch": 0.9790209790209791,
"grad_norm": 0.5200193524360657,
"learning_rate": 1.7886557886557888e-05,
"loss": 0.1952407717704773,
"step": 280
},
{
"epoch": 1.0,
"eval_COMMODITY_f1": 0.4243,
"eval_COMPANY_f1": 0.6975,
"eval_COUNTRY_f1": 0.8103,
"eval_EVENT_f1": 0.0509,
"eval_INFRASTRUCTURE_f1": 0.0265,
"eval_LOCATION_f1": 0.5881,
"eval_MARKET_f1": 0.3048,
"eval_ORGANIZATION_f1": 0.5435,
"eval_PERSON_f1": 0.6313,
"eval_accuracy": 0.9383044022343244,
"eval_f1": 0.5853641863485661,
"eval_loss": 0.18643628060817719,
"eval_precision": 0.5841462582713691,
"eval_recall": 0.5865872037265378,
"eval_runtime": 4.3741,
"eval_samples_per_second": 261.541,
"eval_steps_per_second": 4.115,
"step": 286
},
{
"epoch": 1.048951048951049,
"grad_norm": 0.45493006706237793,
"learning_rate": 1.7575757575757576e-05,
"loss": 0.18901759386062622,
"step": 300
},
{
"epoch": 1.118881118881119,
"grad_norm": 0.4867892861366272,
"learning_rate": 1.7264957264957267e-05,
"loss": 0.18016949892044068,
"step": 320
},
{
"epoch": 1.1888111888111887,
"grad_norm": 0.5517728328704834,
"learning_rate": 1.6954156954156954e-05,
"loss": 0.18962303400039673,
"step": 340
},
{
"epoch": 1.2587412587412588,
"grad_norm": 0.5174335241317749,
"learning_rate": 1.6643356643356645e-05,
"loss": 0.18070975542068482,
"step": 360
},
{
"epoch": 1.3286713286713288,
"grad_norm": 0.5536178350448608,
"learning_rate": 1.6332556332556333e-05,
"loss": 0.17582471370697023,
"step": 380
},
{
"epoch": 1.3986013986013985,
"grad_norm": 0.49411725997924805,
"learning_rate": 1.6021756021756024e-05,
"loss": 0.17660335302352906,
"step": 400
},
{
"epoch": 1.4685314685314685,
"grad_norm": 0.37996432185173035,
"learning_rate": 1.5710955710955715e-05,
"loss": 0.1685216546058655,
"step": 420
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.5632461905479431,
"learning_rate": 1.5400155400155402e-05,
"loss": 0.1708904027938843,
"step": 440
},
{
"epoch": 1.6083916083916083,
"grad_norm": 0.4930890202522278,
"learning_rate": 1.5089355089355091e-05,
"loss": 0.16708383560180665,
"step": 460
},
{
"epoch": 1.6783216783216783,
"grad_norm": 0.5667280554771423,
"learning_rate": 1.4778554778554779e-05,
"loss": 0.16816866397857666,
"step": 480
},
{
"epoch": 1.7482517482517483,
"grad_norm": 0.5394991040229797,
"learning_rate": 1.4467754467754468e-05,
"loss": 0.15965052843093872,
"step": 500
},
{
"epoch": 1.8181818181818183,
"grad_norm": 0.4395284354686737,
"learning_rate": 1.415695415695416e-05,
"loss": 0.16719096899032593,
"step": 520
},
{
"epoch": 1.8881118881118881,
"grad_norm": 0.44819605350494385,
"learning_rate": 1.3846153846153847e-05,
"loss": 0.16433268785476685,
"step": 540
},
{
"epoch": 1.958041958041958,
"grad_norm": 0.630233645439148,
"learning_rate": 1.3535353535353538e-05,
"loss": 0.1701101541519165,
"step": 560
},
{
"epoch": 2.0,
"eval_COMMODITY_f1": 0.5098,
"eval_COMPANY_f1": 0.7415,
"eval_COUNTRY_f1": 0.8307,
"eval_EVENT_f1": 0.2393,
"eval_INFRASTRUCTURE_f1": 0.2385,
"eval_LOCATION_f1": 0.6418,
"eval_MARKET_f1": 0.3517,
"eval_ORGANIZATION_f1": 0.6229,
"eval_PERSON_f1": 0.6953,
"eval_accuracy": 0.9454946483875253,
"eval_f1": 0.6355833913245186,
"eval_loss": 0.15933312475681305,
"eval_precision": 0.6155722446883626,
"eval_recall": 0.6569393067543499,
"eval_runtime": 4.404,
"eval_samples_per_second": 259.765,
"eval_steps_per_second": 4.087,
"step": 572
},
{
"epoch": 2.027972027972028,
"grad_norm": 0.6636055111885071,
"learning_rate": 1.3224553224553225e-05,
"loss": 0.16211626529693604,
"step": 580
},
{
"epoch": 2.097902097902098,
"grad_norm": 0.5422558188438416,
"learning_rate": 1.2913752913752915e-05,
"loss": 0.1504289388656616,
"step": 600
},
{
"epoch": 2.167832167832168,
"grad_norm": 0.5456656813621521,
"learning_rate": 1.2602952602952606e-05,
"loss": 0.15346094369888305,
"step": 620
},
{
"epoch": 2.237762237762238,
"grad_norm": 0.5587140321731567,
"learning_rate": 1.2292152292152293e-05,
"loss": 0.15093343257904052,
"step": 640
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.4914584457874298,
"learning_rate": 1.1981351981351982e-05,
"loss": 0.14887770414352416,
"step": 660
},
{
"epoch": 2.3776223776223775,
"grad_norm": 0.42390987277030945,
"learning_rate": 1.1670551670551672e-05,
"loss": 0.15100462436676027,
"step": 680
},
{
"epoch": 2.4475524475524475,
"grad_norm": 0.42798611521720886,
"learning_rate": 1.1359751359751361e-05,
"loss": 0.14843168258666992,
"step": 700
},
{
"epoch": 2.5174825174825175,
"grad_norm": 0.559911847114563,
"learning_rate": 1.1048951048951048e-05,
"loss": 0.14720556735992432,
"step": 720
},
{
"epoch": 2.5874125874125875,
"grad_norm": 0.5178841948509216,
"learning_rate": 1.073815073815074e-05,
"loss": 0.14619035720825196,
"step": 740
},
{
"epoch": 2.6573426573426575,
"grad_norm": 0.6083295941352844,
"learning_rate": 1.0427350427350429e-05,
"loss": 0.14750727415084838,
"step": 760
},
{
"epoch": 2.7272727272727275,
"grad_norm": 0.46528416872024536,
"learning_rate": 1.0116550116550116e-05,
"loss": 0.14334226846694947,
"step": 780
},
{
"epoch": 2.797202797202797,
"grad_norm": 0.48938772082328796,
"learning_rate": 9.805749805749807e-06,
"loss": 0.14555764198303223,
"step": 800
},
{
"epoch": 2.867132867132867,
"grad_norm": 0.4275546371936798,
"learning_rate": 9.494949494949497e-06,
"loss": 0.14947701692581178,
"step": 820
},
{
"epoch": 2.937062937062937,
"grad_norm": 0.5199507474899292,
"learning_rate": 9.184149184149184e-06,
"loss": 0.14035249948501588,
"step": 840
},
{
"epoch": 3.0,
"eval_COMMODITY_f1": 0.4887,
"eval_COMPANY_f1": 0.7755,
"eval_COUNTRY_f1": 0.8455,
"eval_EVENT_f1": 0.3252,
"eval_INFRASTRUCTURE_f1": 0.3,
"eval_LOCATION_f1": 0.6741,
"eval_MARKET_f1": 0.4189,
"eval_ORGANIZATION_f1": 0.6693,
"eval_PERSON_f1": 0.7555,
"eval_accuracy": 0.9481756978768325,
"eval_f1": 0.6681568653232964,
"eval_loss": 0.15031008422374725,
"eval_precision": 0.6422928648170385,
"eval_recall": 0.6961912590765859,
"eval_runtime": 4.4312,
"eval_samples_per_second": 258.167,
"eval_steps_per_second": 4.062,
"step": 858
},
{
"epoch": 3.006993006993007,
"grad_norm": 0.5143063068389893,
"learning_rate": 8.873348873348873e-06,
"loss": 0.13901138305664062,
"step": 860
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.46927395462989807,
"learning_rate": 8.562548562548563e-06,
"loss": 0.13868144750595093,
"step": 880
},
{
"epoch": 3.1468531468531467,
"grad_norm": 0.4429858922958374,
"learning_rate": 8.251748251748254e-06,
"loss": 0.13494281768798827,
"step": 900
},
{
"epoch": 3.2167832167832167,
"grad_norm": 0.4369621276855469,
"learning_rate": 7.940947940947941e-06,
"loss": 0.13719457387924194,
"step": 920
},
{
"epoch": 3.2867132867132867,
"grad_norm": 0.5295616388320923,
"learning_rate": 7.63014763014763e-06,
"loss": 0.13230640888214112,
"step": 940
},
{
"epoch": 3.3566433566433567,
"grad_norm": 0.7585546374320984,
"learning_rate": 7.31934731934732e-06,
"loss": 0.13142707347869872,
"step": 960
},
{
"epoch": 3.4265734265734267,
"grad_norm": 0.7057129740715027,
"learning_rate": 7.008547008547009e-06,
"loss": 0.13272271156311036,
"step": 980
},
{
"epoch": 3.4965034965034967,
"grad_norm": 0.4822175204753876,
"learning_rate": 6.697746697746699e-06,
"loss": 0.13051105737686158,
"step": 1000
},
{
"epoch": 3.5664335664335667,
"grad_norm": 0.6353693008422852,
"learning_rate": 6.3869463869463875e-06,
"loss": 0.1364367723464966,
"step": 1020
},
{
"epoch": 3.6363636363636362,
"grad_norm": 0.6322542428970337,
"learning_rate": 6.076146076146077e-06,
"loss": 0.13394793272018432,
"step": 1040
},
{
"epoch": 3.7062937062937062,
"grad_norm": 0.6142822504043579,
"learning_rate": 5.765345765345766e-06,
"loss": 0.13053072690963746,
"step": 1060
},
{
"epoch": 3.7762237762237763,
"grad_norm": 0.4618857800960541,
"learning_rate": 5.4545454545454545e-06,
"loss": 0.13272134065628052,
"step": 1080
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.5411708354949951,
"learning_rate": 5.1437451437451446e-06,
"loss": 0.1273587703704834,
"step": 1100
},
{
"epoch": 3.916083916083916,
"grad_norm": 0.438820481300354,
"learning_rate": 4.832944832944833e-06,
"loss": 0.13514026403427123,
"step": 1120
},
{
"epoch": 3.986013986013986,
"grad_norm": 0.580314040184021,
"learning_rate": 4.522144522144522e-06,
"loss": 0.13724528551101683,
"step": 1140
},
{
"epoch": 4.0,
"eval_COMMODITY_f1": 0.5015,
"eval_COMPANY_f1": 0.7911,
"eval_COUNTRY_f1": 0.8487,
"eval_EVENT_f1": 0.3324,
"eval_INFRASTRUCTURE_f1": 0.3305,
"eval_LOCATION_f1": 0.6716,
"eval_MARKET_f1": 0.4217,
"eval_ORGANIZATION_f1": 0.68,
"eval_PERSON_f1": 0.7632,
"eval_accuracy": 0.9506155577771408,
"eval_f1": 0.6738893476465732,
"eval_loss": 0.1432034969329834,
"eval_precision": 0.6506984754736238,
"eval_recall": 0.6987943553911494,
"eval_runtime": 4.3982,
"eval_samples_per_second": 260.109,
"eval_steps_per_second": 4.093,
"step": 1144
}
],
"logging_steps": 20,
"max_steps": 1430,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2392209436311552.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}