kanak8278's picture
End of training
a26b81d
raw
history blame contribute delete
210 Bytes
{
"epoch": 3.0,
"total_flos": 1.2614074495819776e+18,
"train_loss": 0.17057580525059088,
"train_runtime": 1281.8017,
"train_samples_per_second": 39.217,
"train_steps_per_second": 1.224
}