josecannete's picture
adding model fine-tuned on QA (SQAC)
cd6c5a7
{
"best_metric": 79.90477476395145,
"best_model_checkpoint": "/data/jcanete/all_results/sqac/albeto_base/epochs_3_bs_16_lr_5e-5/checkpoint-2000",
"epoch": 3.0,
"global_step": 3114,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.19,
"eval_exact_match": 40.55793991416309,
"eval_f1": 61.9475026058651,
"step": 200
},
{
"epoch": 0.39,
"eval_exact_match": 50.16094420600859,
"eval_f1": 70.94487746222757,
"step": 400
},
{
"epoch": 0.48,
"learning_rate": 4.206807964033398e-05,
"loss": 1.908,
"step": 500
},
{
"epoch": 0.58,
"eval_exact_match": 54.77467811158798,
"eval_f1": 73.11200761877414,
"step": 600
},
{
"epoch": 0.77,
"eval_exact_match": 55.90128755364807,
"eval_f1": 74.35378026715777,
"step": 800
},
{
"epoch": 0.96,
"learning_rate": 3.405587668593449e-05,
"loss": 1.1479,
"step": 1000
},
{
"epoch": 0.96,
"eval_exact_match": 58.74463519313305,
"eval_f1": 77.15822408085789,
"step": 1000
},
{
"epoch": 1.16,
"eval_exact_match": 58.90557939914163,
"eval_f1": 76.90716161104929,
"step": 1200
},
{
"epoch": 1.35,
"eval_exact_match": 59.49570815450644,
"eval_f1": 77.54387552767798,
"step": 1400
},
{
"epoch": 1.45,
"learning_rate": 2.6027617212588313e-05,
"loss": 0.7268,
"step": 1500
},
{
"epoch": 1.54,
"eval_exact_match": 59.227467811158796,
"eval_f1": 77.8200127931914,
"step": 1600
},
{
"epoch": 1.73,
"eval_exact_match": 61.31974248927039,
"eval_f1": 78.96026149786353,
"step": 1800
},
{
"epoch": 1.93,
"learning_rate": 1.7999357739242135e-05,
"loss": 0.683,
"step": 2000
},
{
"epoch": 1.93,
"eval_exact_match": 62.71459227467811,
"eval_f1": 79.90477476395145,
"step": 2000
},
{
"epoch": 2.12,
"eval_exact_match": 62.01716738197425,
"eval_f1": 79.35780895037563,
"step": 2200
},
{
"epoch": 2.31,
"eval_exact_match": 62.23175965665236,
"eval_f1": 79.16073100037653,
"step": 2400
},
{
"epoch": 2.41,
"learning_rate": 9.971098265895954e-06,
"loss": 0.4189,
"step": 2500
},
{
"epoch": 2.5,
"eval_exact_match": 60.89055793991416,
"eval_f1": 79.00549485950138,
"step": 2600
},
{
"epoch": 2.7,
"eval_exact_match": 61.85622317596567,
"eval_f1": 79.2496307825153,
"step": 2800
},
{
"epoch": 2.89,
"learning_rate": 1.9428387925497756e-06,
"loss": 0.3417,
"step": 3000
},
{
"epoch": 2.89,
"eval_exact_match": 62.55364806866953,
"eval_f1": 79.81438829072098,
"step": 3000
},
{
"epoch": 3.0,
"step": 3114,
"total_flos": 1074824219579904.0,
"train_loss": 0.8525694306262732,
"train_runtime": 946.0422,
"train_samples_per_second": 52.634,
"train_steps_per_second": 3.292
}
],
"max_steps": 3114,
"num_train_epochs": 3,
"total_flos": 1074824219579904.0,
"trial_name": null,
"trial_params": null
}