xlm-roberta-large-lemma-eu / trainer_state.json
oliat's picture
uploading files eu model
c86a048 verified
raw
history blame
10.4 kB
{
"best_metric": 0.9623383341955509,
"best_model_checkpoint": ".//debugged_eu_bdt_ses_udpipe_16_0.01_0.00005_20_04-23-24_02-43/checkpoint-8550",
"epoch": 20.0,
"global_step": 9000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"learning_rate": 4.166666666666667e-05,
"loss": 2.3424,
"step": 450
},
{
"epoch": 1.0,
"eval_accuracy": 0.8618712108798952,
"eval_f1": 0.8046604387530177,
"eval_loss": 0.7042863368988037,
"eval_precision": 0.8160527996593571,
"eval_recall": 0.7935817805383023,
"eval_runtime": 5.6554,
"eval_samples_per_second": 159.141,
"eval_steps_per_second": 19.981,
"step": 450
},
{
"epoch": 2.0,
"learning_rate": 4.787234042553192e-05,
"loss": 0.5837,
"step": 900
},
{
"epoch": 2.0,
"eval_accuracy": 0.9321645092577421,
"eval_f1": 0.9056838826888141,
"eval_loss": 0.37773051857948303,
"eval_precision": 0.9082769390942218,
"eval_recall": 0.9031055900621118,
"eval_runtime": 4.8592,
"eval_samples_per_second": 185.216,
"eval_steps_per_second": 23.255,
"step": 900
},
{
"epoch": 3.0,
"learning_rate": 4.5212765957446815e-05,
"loss": 0.3375,
"step": 1350
},
{
"epoch": 3.0,
"eval_accuracy": 0.9487956742585614,
"eval_f1": 0.9285084763336616,
"eval_loss": 0.293639600276947,
"eval_precision": 0.9300031155883269,
"eval_recall": 0.9270186335403726,
"eval_runtime": 4.8723,
"eval_samples_per_second": 184.719,
"eval_steps_per_second": 23.192,
"step": 1350
},
{
"epoch": 4.0,
"learning_rate": 4.2553191489361704e-05,
"loss": 0.2443,
"step": 1800
},
{
"epoch": 4.0,
"eval_accuracy": 0.9546944125839751,
"eval_f1": 0.9368006630750103,
"eval_loss": 0.2582251727581024,
"eval_precision": 0.9375777685607631,
"eval_recall": 0.9360248447204969,
"eval_runtime": 4.875,
"eval_samples_per_second": 184.617,
"eval_steps_per_second": 23.18,
"step": 1800
},
{
"epoch": 5.0,
"learning_rate": 3.9893617021276594e-05,
"loss": 0.1709,
"step": 2250
},
{
"epoch": 5.0,
"eval_accuracy": 0.9590365394068491,
"eval_f1": 0.9431171786120591,
"eval_loss": 0.23951154947280884,
"eval_precision": 0.9422401322587312,
"eval_recall": 0.9439958592132505,
"eval_runtime": 4.8802,
"eval_samples_per_second": 184.419,
"eval_steps_per_second": 23.155,
"step": 2250
},
{
"epoch": 6.0,
"learning_rate": 3.723404255319149e-05,
"loss": 0.1282,
"step": 2700
},
{
"epoch": 6.0,
"eval_accuracy": 0.9626413239390463,
"eval_f1": 0.9479349963771866,
"eval_loss": 0.2380053848028183,
"eval_precision": 0.9478368867729249,
"eval_recall": 0.9480331262939958,
"eval_runtime": 4.8713,
"eval_samples_per_second": 184.757,
"eval_steps_per_second": 23.197,
"step": 2700
},
{
"epoch": 7.0,
"learning_rate": 3.4574468085106386e-05,
"loss": 0.0975,
"step": 3150
},
{
"epoch": 7.0,
"eval_accuracy": 0.9656726200229395,
"eval_f1": 0.9517748111352583,
"eval_loss": 0.23213887214660645,
"eval_precision": 0.9514794123732672,
"eval_recall": 0.9520703933747412,
"eval_runtime": 5.071,
"eval_samples_per_second": 177.48,
"eval_steps_per_second": 22.284,
"step": 3150
},
{
"epoch": 8.0,
"learning_rate": 3.191489361702128e-05,
"loss": 0.0721,
"step": 3600
},
{
"epoch": 8.0,
"eval_accuracy": 0.9674750122890382,
"eval_f1": 0.9542747656775931,
"eval_loss": 0.2307664304971695,
"eval_precision": 0.9547197181639209,
"eval_recall": 0.9538302277432712,
"eval_runtime": 4.872,
"eval_samples_per_second": 184.728,
"eval_steps_per_second": 23.194,
"step": 3600
},
{
"epoch": 9.0,
"learning_rate": 2.925531914893617e-05,
"loss": 0.0561,
"step": 4050
},
{
"epoch": 9.0,
"eval_accuracy": 0.9686219891856463,
"eval_f1": 0.9559272877932571,
"eval_loss": 0.2352980375289917,
"eval_precision": 0.9564721732822055,
"eval_recall": 0.9553830227743271,
"eval_runtime": 4.8623,
"eval_samples_per_second": 185.097,
"eval_steps_per_second": 23.24,
"step": 4050
},
{
"epoch": 10.0,
"learning_rate": 2.6595744680851064e-05,
"loss": 0.0395,
"step": 4500
},
{
"epoch": 10.0,
"eval_accuracy": 0.9672292315254792,
"eval_f1": 0.954816003312458,
"eval_loss": 0.24887719750404358,
"eval_precision": 0.9547665873098024,
"eval_recall": 0.9548654244306418,
"eval_runtime": 4.8696,
"eval_samples_per_second": 184.821,
"eval_steps_per_second": 23.205,
"step": 4500
},
{
"epoch": 11.0,
"learning_rate": 2.393617021276596e-05,
"loss": 0.0284,
"step": 4950
},
{
"epoch": 11.0,
"eval_accuracy": 0.9688677699492053,
"eval_f1": 0.9560450925638639,
"eval_loss": 0.2504814863204956,
"eval_precision": 0.955156023971895,
"eval_recall": 0.9569358178053831,
"eval_runtime": 5.8981,
"eval_samples_per_second": 152.592,
"eval_steps_per_second": 19.159,
"step": 4950
},
{
"epoch": 12.0,
"learning_rate": 2.1276595744680852e-05,
"loss": 0.0178,
"step": 5400
},
{
"epoch": 12.0,
"eval_accuracy": 0.9695231853186957,
"eval_f1": 0.9567693502459229,
"eval_loss": 0.2542861998081207,
"eval_precision": 0.9570170895908856,
"eval_recall": 0.9565217391304348,
"eval_runtime": 5.0143,
"eval_samples_per_second": 179.488,
"eval_steps_per_second": 22.536,
"step": 5400
},
{
"epoch": 13.0,
"learning_rate": 1.8617021276595745e-05,
"loss": 0.0119,
"step": 5850
},
{
"epoch": 13.0,
"eval_accuracy": 0.9688677699492053,
"eval_f1": 0.9562247749146229,
"eval_loss": 0.2555958926677704,
"eval_precision": 0.9559279950341403,
"eval_recall": 0.9565217391304348,
"eval_runtime": 5.0192,
"eval_samples_per_second": 179.312,
"eval_steps_per_second": 22.514,
"step": 5850
},
{
"epoch": 14.0,
"learning_rate": 1.595744680851064e-05,
"loss": 0.008,
"step": 6300
},
{
"epoch": 14.0,
"eval_accuracy": 0.9707520891364902,
"eval_f1": 0.9594783418723801,
"eval_loss": 0.26319199800491333,
"eval_precision": 0.9593294008072027,
"eval_recall": 0.9596273291925466,
"eval_runtime": 5.2921,
"eval_samples_per_second": 170.065,
"eval_steps_per_second": 21.353,
"step": 6300
},
{
"epoch": 15.0,
"learning_rate": 1.3297872340425532e-05,
"loss": 0.0068,
"step": 6750
},
{
"epoch": 15.0,
"eval_accuracy": 0.9711617237424217,
"eval_f1": 0.9603186921206478,
"eval_loss": 0.260258287191391,
"eval_precision": 0.9598717550935981,
"eval_recall": 0.9607660455486542,
"eval_runtime": 5.016,
"eval_samples_per_second": 179.425,
"eval_steps_per_second": 22.528,
"step": 6750
},
{
"epoch": 16.0,
"learning_rate": 1.0638297872340426e-05,
"loss": 0.0044,
"step": 7200
},
{
"epoch": 16.0,
"eval_accuracy": 0.9702605276093724,
"eval_f1": 0.9582665356570306,
"eval_loss": 0.27104565501213074,
"eval_precision": 0.9574248217422755,
"eval_recall": 0.9591097308488613,
"eval_runtime": 5.1972,
"eval_samples_per_second": 173.172,
"eval_steps_per_second": 21.743,
"step": 7200
},
{
"epoch": 17.0,
"learning_rate": 7.97872340425532e-06,
"loss": 0.0032,
"step": 7650
},
{
"epoch": 17.0,
"eval_accuracy": 0.9718171391119121,
"eval_f1": 0.9608736155677465,
"eval_loss": 0.27112799882888794,
"eval_precision": 0.9607741668391637,
"eval_recall": 0.9609730848861283,
"eval_runtime": 5.352,
"eval_samples_per_second": 168.161,
"eval_steps_per_second": 21.114,
"step": 7650
},
{
"epoch": 18.0,
"learning_rate": 5.319148936170213e-06,
"loss": 0.0024,
"step": 8100
},
{
"epoch": 18.0,
"eval_accuracy": 0.9717352121907259,
"eval_f1": 0.9607863424728401,
"eval_loss": 0.27230679988861084,
"eval_precision": 0.9602895553257498,
"eval_recall": 0.9612836438923396,
"eval_runtime": 5.124,
"eval_samples_per_second": 175.645,
"eval_steps_per_second": 22.053,
"step": 8100
},
{
"epoch": 19.0,
"learning_rate": 2.6595744680851065e-06,
"loss": 0.0019,
"step": 8550
},
{
"epoch": 19.0,
"eval_accuracy": 0.9726364083237752,
"eval_f1": 0.9623383341955509,
"eval_loss": 0.27249038219451904,
"eval_precision": 0.9618407445708377,
"eval_recall": 0.9628364389233954,
"eval_runtime": 5.0361,
"eval_samples_per_second": 178.708,
"eval_steps_per_second": 22.438,
"step": 8550
},
{
"epoch": 20.0,
"learning_rate": 0.0,
"loss": 0.0015,
"step": 9000
},
{
"epoch": 20.0,
"eval_accuracy": 0.9720629198754711,
"eval_f1": 0.9611961920529801,
"eval_loss": 0.27408093214035034,
"eval_precision": 0.9607985105502689,
"eval_recall": 0.9615942028985507,
"eval_runtime": 5.1327,
"eval_samples_per_second": 175.347,
"eval_steps_per_second": 22.016,
"step": 9000
},
{
"epoch": 20.0,
"step": 9000,
"total_flos": 1.1952384086166018e+16,
"train_loss": 0.20792188170221118,
"train_runtime": 4350.4019,
"train_samples_per_second": 33.077,
"train_steps_per_second": 2.069
}
],
"max_steps": 9000,
"num_train_epochs": 20,
"total_flos": 1.1952384086166018e+16,
"trial_name": null,
"trial_params": null
}