josecannete's picture
adding model finetuned on tar
fe6b61c
{
"best_metric": 77.32269615784104,
"best_model_checkpoint": "/home/jcanete/ft-data/all_results/tar/roberta_large_bne/epochs_2_bs_32_lr_2e-5/checkpoint-4500",
"epoch": 2.0,
"global_step": 5486,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.11,
"eval_exact_match": 52.72469252601703,
"eval_f1": 70.31979377504159,
"step": 300
},
{
"epoch": 0.18,
"learning_rate": 1.8177178271965002e-05,
"loss": 2.0124,
"step": 500
},
{
"epoch": 0.22,
"eval_exact_match": 55.37369914853358,
"eval_f1": 72.79425345069818,
"step": 600
},
{
"epoch": 0.33,
"eval_exact_match": 57.142857142857146,
"eval_f1": 74.00444934296571,
"step": 900
},
{
"epoch": 0.36,
"learning_rate": 1.6354356543930007e-05,
"loss": 1.6158,
"step": 1000
},
{
"epoch": 0.44,
"eval_exact_match": 57.899716177861876,
"eval_f1": 75.07716154126113,
"step": 1200
},
{
"epoch": 0.55,
"learning_rate": 1.4531534815895008e-05,
"loss": 1.5421,
"step": 1500
},
{
"epoch": 0.55,
"eval_exact_match": 59.08230842005676,
"eval_f1": 75.72655876274406,
"step": 1500
},
{
"epoch": 0.66,
"eval_exact_match": 58.77010406811731,
"eval_f1": 75.5986676698771,
"step": 1800
},
{
"epoch": 0.73,
"learning_rate": 1.2712358731316079e-05,
"loss": 1.4971,
"step": 2000
},
{
"epoch": 0.77,
"eval_exact_match": 59.40397350993378,
"eval_f1": 76.56550651885689,
"step": 2100
},
{
"epoch": 0.87,
"eval_exact_match": 60.33112582781457,
"eval_f1": 77.04475615318619,
"step": 2400
},
{
"epoch": 0.91,
"learning_rate": 1.0889537003281081e-05,
"loss": 1.4532,
"step": 2500
},
{
"epoch": 0.98,
"eval_exact_match": 60.179754020813625,
"eval_f1": 76.98820150102628,
"step": 2700
},
{
"epoch": 1.09,
"learning_rate": 9.066715275246082e-06,
"loss": 1.1553,
"step": 3000
},
{
"epoch": 1.09,
"eval_exact_match": 59.735099337748345,
"eval_f1": 76.60782983695076,
"step": 3000
},
{
"epoch": 1.2,
"eval_exact_match": 59.7918637653737,
"eval_f1": 76.76732218063003,
"step": 3300
},
{
"epoch": 1.28,
"learning_rate": 7.243893547211084e-06,
"loss": 0.9124,
"step": 3500
},
{
"epoch": 1.31,
"eval_exact_match": 59.83916745506149,
"eval_f1": 76.96753684496848,
"step": 3600
},
{
"epoch": 1.42,
"eval_exact_match": 59.99053926206244,
"eval_f1": 76.84857447005851,
"step": 3900
},
{
"epoch": 1.46,
"learning_rate": 5.421071819176085e-06,
"loss": 0.8835,
"step": 4000
},
{
"epoch": 1.53,
"eval_exact_match": 59.422894985808895,
"eval_f1": 76.94657837246879,
"step": 4200
},
{
"epoch": 1.64,
"learning_rate": 3.6018957345971567e-06,
"loss": 0.8727,
"step": 4500
},
{
"epoch": 1.64,
"eval_exact_match": 60.416272469252604,
"eval_f1": 77.32269615784104,
"step": 4500
},
{
"epoch": 1.75,
"eval_exact_match": 60.2081362346263,
"eval_f1": 77.1919063305815,
"step": 4800
},
{
"epoch": 1.82,
"learning_rate": 1.7827196500182284e-06,
"loss": 0.8503,
"step": 5000
},
{
"epoch": 1.86,
"eval_exact_match": 60.06622516556291,
"eval_f1": 77.17743863949298,
"step": 5100
},
{
"epoch": 1.97,
"eval_exact_match": 60.31220435193945,
"eval_f1": 77.30000861688039,
"step": 5400
},
{
"epoch": 2.0,
"step": 5486,
"total_flos": 1.1863529489772403e+17,
"train_loss": 1.2413271612987178,
"train_runtime": 3098.796,
"train_samples_per_second": 56.634,
"train_steps_per_second": 1.77
}
],
"max_steps": 5486,
"num_train_epochs": 2,
"total_flos": 1.1863529489772403e+17,
"trial_name": null,
"trial_params": null
}