ALM-AHME's picture
End of training
d863a9d
raw
history blame
206 Bytes
{
"epoch": 11.95,
"total_flos": 7.740923166391597e+18,
"train_loss": 0.8764253104900757,
"train_runtime": 6223.024,
"train_samples_per_second": 7.04,
"train_steps_per_second": 0.22
}