distilled-mt5-small-0.005-0.25 / train_results.json
Lvxue's picture
End of training
1bdfc49
{
"epoch": 5.0,
"train_loss": 3.305007841796875,
"train_runtime": 2877.6666,
"train_samples": 10000,
"train_samples_per_second": 17.375,
"train_steps_per_second": 4.344
}