djbp's picture
End of training
25f05a6 verified
raw
history blame
203 Bytes
{
"epoch": 7.0,
"total_flos": 8226380544399360.0,
"train_loss": 0.14502547468457902,
"train_runtime": 156.7549,
"train_samples_per_second": 0.67,
"train_steps_per_second": 0.045
}