paul
End of training
dcc41cf
{
"epoch": 9.99,
"total_flos": 4.005358668612661e+18,
"train_loss": 0.31261438488960264,
"train_runtime": 5959.0067,
"train_samples_per_second": 8.683,
"train_steps_per_second": 0.067
}