distilled-mt5-small-b0.03 / train_results.json
Lvxue's picture
End of training
4f02add
raw
history blame contribute delete
194 Bytes
{
"epoch": 5.0,
"train_loss": 3.976136337890625,
"train_runtime": 2627.9564,
"train_samples": 10000,
"train_samples_per_second": 19.026,
"train_steps_per_second": 4.757
}