gruhntm's picture
End of training
a8192ce verified
raw
history blame contribute delete
209 Bytes
{
"epoch": 3.0,
"total_flos": 1.4499253204608614e+18,
"train_loss": 0.36645709030460893,
"train_runtime": 903.5275,
"train_samples_per_second": 64.547,
"train_steps_per_second": 0.505
}