aricibo's picture
End of training
611dcdf
{
"epoch": 3.0,
"total_flos": 1.8299091784499528e+18,
"train_loss": 0.1583421324428759,
"train_runtime": 2450.017,
"train_samples_per_second": 29.755,
"train_steps_per_second": 0.233
}