t1msan's picture
End of training
dfa127e verified
{
"epoch": 5.0,
"total_flos": 5.914482579184435e+18,
"train_loss": 0.04452576920570385,
"train_runtime": 5366.5614,
"train_samples_per_second": 44.339,
"train_steps_per_second": 0.231
}