esfrankel17's picture
End of training
ccc134a verified
raw
history blame contribute delete
206 Bytes
{
"epoch": 1.6,
"total_flos": 1.922102501100749e+16,
"train_loss": 1.818156321843465,
"train_runtime": 1285.2745,
"train_samples_per_second": 4.479,
"train_steps_per_second": 0.002
}