djbp's picture
End of training
9ad18f7 verified
raw
history blame
221 Bytes
{
"epoch": 6.885245901639344,
"total_flos": 4.180411553219629e+18,
"train_loss": 0.4180879854020618,
"train_runtime": 6613.2188,
"train_samples_per_second": 8.194,
"train_steps_per_second": 0.016
}