phunc20's picture
End of training
d98b3dd
raw
history blame
208 Bytes
{
"epoch": 20.0,
"total_flos": 1.7747176135901184e+17,
"train_loss": 0.1779410476485888,
"train_runtime": 979.9272,
"train_samples_per_second": 7.286,
"train_steps_per_second": 0.061
}