djbp's picture
End of training
611a9b2 verified
raw
history blame contribute delete
223 Bytes
{
"epoch": 6.885245901639344,
"total_flos": 4.1785312376666235e+18,
"train_loss": 0.3947783969697498,
"train_runtime": 11418.3292,
"train_samples_per_second": 4.743,
"train_steps_per_second": 0.009
}