chchen's picture
End of training
2c092db verified
{
"epoch": 4.994228549442093,
"total_flos": 8.1511813636463e+17,
"train_loss": 0.041532339387284865,
"train_runtime": 35427.3978,
"train_samples_per_second": 1.467,
"train_steps_per_second": 0.092
}