Jzuluaga's picture
updating the repo with the fine-tuned model
03154fa
raw
history blame
400 Bytes
{
"epoch": 21.23,
"eval_loss": 0.84695965051651,
"eval_runtime": 90.2768,
"eval_samples": 2822,
"eval_samples_per_second": 31.259,
"eval_steps_per_second": 2.614,
"eval_wer": 0.18976083707025412,
"train_loss": 0.6778903991699219,
"train_runtime": 11598.0436,
"train_samples": 11291,
"train_samples_per_second": 20.693,
"train_steps_per_second": 0.862
}