distilled-mt5-small-0.05-0.25 / train_results.json
Lvxue's picture
End of training
58fb908
{
"epoch": 5.0,
"train_loss": 4.38795890625,
"train_runtime": 3403.5404,
"train_samples": 10000,
"train_samples_per_second": 14.691,
"train_steps_per_second": 3.673
}