rob-base-gc1 / trainer_state.json
nbroad's picture
nbroad HF staff
End of training
b0116e4
raw
history blame
4.64 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"global_step": 2482,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.08,
"learning_rate": 4.0160642570281125e-05,
"loss": 3.6101,
"step": 100
},
{
"epoch": 0.16,
"learning_rate": 8.032128514056225e-05,
"loss": 1.5644,
"step": 200
},
{
"epoch": 0.24,
"learning_rate": 9.771607702642186e-05,
"loss": 1.4534,
"step": 300
},
{
"epoch": 0.32,
"learning_rate": 9.323779668607256e-05,
"loss": 1.2308,
"step": 400
},
{
"epoch": 0.4,
"learning_rate": 8.875951634572325e-05,
"loss": 1.0461,
"step": 500
},
{
"epoch": 0.48,
"learning_rate": 8.428123600537394e-05,
"loss": 1.2082,
"step": 600
},
{
"epoch": 0.56,
"learning_rate": 7.980295566502463e-05,
"loss": 1.1456,
"step": 700
},
{
"epoch": 0.64,
"learning_rate": 7.532467532467533e-05,
"loss": 1.0767,
"step": 800
},
{
"epoch": 0.73,
"learning_rate": 7.084639498432603e-05,
"loss": 1.136,
"step": 900
},
{
"epoch": 0.81,
"learning_rate": 6.636811464397672e-05,
"loss": 1.041,
"step": 1000
},
{
"epoch": 0.89,
"learning_rate": 6.188983430362741e-05,
"loss": 1.1013,
"step": 1100
},
{
"epoch": 0.97,
"learning_rate": 5.741155396327811e-05,
"loss": 0.9227,
"step": 1200
},
{
"epoch": 1.0,
"eval_HasAns_exact": 68.0616620528687,
"eval_HasAns_f1": 74.92241418708792,
"eval_HasAns_total": 18423,
"eval_NoAns_exact": 72.74377165484243,
"eval_NoAns_f1": 72.74377165484243,
"eval_NoAns_total": 6061,
"eval_best_exact": 69.22071556935141,
"eval_best_exact_thresh": 0.0,
"eval_best_f1": 74.3830924917785,
"eval_best_f1_thresh": 0.0,
"eval_exact": 69.22071556935141,
"eval_f1": 74.38309249177856,
"eval_total": 24484,
"step": 1241
},
{
"epoch": 1.05,
"learning_rate": 5.29332736229288e-05,
"loss": 0.9157,
"step": 1300
},
{
"epoch": 1.13,
"learning_rate": 4.845499328257949e-05,
"loss": 0.7482,
"step": 1400
},
{
"epoch": 1.21,
"learning_rate": 4.397671294223019e-05,
"loss": 0.7523,
"step": 1500
},
{
"epoch": 1.29,
"learning_rate": 3.949843260188088e-05,
"loss": 0.8341,
"step": 1600
},
{
"epoch": 1.37,
"learning_rate": 3.5020152261531576e-05,
"loss": 0.7643,
"step": 1700
},
{
"epoch": 1.45,
"learning_rate": 3.0541871921182266e-05,
"loss": 0.7429,
"step": 1800
},
{
"epoch": 1.53,
"learning_rate": 2.6063591580832963e-05,
"loss": 0.7005,
"step": 1900
},
{
"epoch": 1.61,
"learning_rate": 2.1585311240483657e-05,
"loss": 0.7959,
"step": 2000
},
{
"epoch": 1.69,
"learning_rate": 1.710703090013435e-05,
"loss": 0.838,
"step": 2100
},
{
"epoch": 1.77,
"learning_rate": 1.2628750559785044e-05,
"loss": 0.6559,
"step": 2200
},
{
"epoch": 1.85,
"learning_rate": 8.150470219435737e-06,
"loss": 0.7658,
"step": 2300
},
{
"epoch": 1.93,
"learning_rate": 3.672189879086431e-06,
"loss": 0.7433,
"step": 2400
},
{
"epoch": 2.0,
"eval_HasAns_exact": 67.86625413884818,
"eval_HasAns_f1": 74.35217861420978,
"eval_HasAns_total": 18423,
"eval_NoAns_exact": 78.12242204256724,
"eval_NoAns_f1": 78.12242204256724,
"eval_NoAns_total": 6061,
"eval_best_exact": 70.40516255513805,
"eval_best_exact_thresh": 0.0,
"eval_best_f1": 75.28550018826877,
"eval_best_f1_thresh": 0.0,
"eval_exact": 70.40516255513805,
"eval_f1": 75.28550018826883,
"eval_total": 24484,
"step": 2482
},
{
"epoch": 2.0,
"step": 2482,
"total_flos": 7.969241705088025e+18,
"train_loss": 1.0671158636125415,
"train_runtime": 2476.3519,
"train_samples_per_second": 256.584,
"train_steps_per_second": 1.002
}
],
"max_steps": 2482,
"num_train_epochs": 2,
"start_time": 1661273831.2880511,
"total_flos": 7.969241705088025e+18,
"trial_name": null,
"trial_params": null
}