Jzuluaga's picture
updating the repo with the fine-tuned model
a6940a5
{
"epoch": 256.41,
"eval_loss": 0.09882301837205887,
"eval_runtime": 66.9757,
"eval_samples": 1857,
"eval_samples_per_second": 27.726,
"eval_steps_per_second": 1.165,
"eval_wer": 0.07359326539654408
}