weight_decay_02 / all_results.json
sedrickkeh's picture
End of training
c9a2c23 verified
{
"epoch": 2.9878213802435725,
"total_flos": 7.10230108051551e+18,
"train_loss": 0.4666922677686249,
"train_runtime": 60539.5618,
"train_samples_per_second": 2.343,
"train_steps_per_second": 0.005
}