chchen's picture
End of training
bbe930d verified
raw
history blame contribute delete
223 Bytes
{
"epoch": 4.938271604938271,
"total_flos": 1.5526893871949414e+17,
"train_loss": 0.07336431298404932,
"train_runtime": 3965.4921,
"train_samples_per_second": 0.511,
"train_steps_per_second": 0.032
}