distilled-mt5-small-0.005-1 / train_results.json
Lvxue's picture
End of training
cec6325
{
"epoch": 5.0,
"train_loss": 3.999654716796875,
"train_runtime": 2751.7981,
"train_samples": 10000,
"train_samples_per_second": 18.17,
"train_steps_per_second": 4.542
}