vincentiussgk's picture
End of training
80a655d
raw
history blame
206 Bytes
{
"epoch": 2.98,
"total_flos": 1.0405550045270016e+18,
"train_loss": 2.129908829643613,
"train_runtime": 618.168,
"train_samples_per_second": 21.839,
"train_steps_per_second": 0.17
}