wikitext_roberta-base / train_results.json
gary109's picture
End of training
4dc1362
raw
history blame
196 Bytes
{
"epoch": 19.99,
"train_loss": 1.3203882475157043,
"train_runtime": 9311.1992,
"train_samples": 4798,
"train_samples_per_second": 10.306,
"train_steps_per_second": 0.079
}