tea-cup's picture
End of training
551608b
raw
history blame
208 Bytes
{
"epoch": 2.99,
"total_flos": 1.2596143944940093e+18,
"train_loss": 0.304791094678821,
"train_runtime": 5290.6872,
"train_samples_per_second": 9.611,
"train_steps_per_second": 0.075
}