gpt2-larger-walser / train_results.json
Jonas
walser larger commit
01c3411
{
"epoch": 5.0,
"train_loss": 4.937346829501065,
"train_runtime": 481.6631,
"train_samples": 1755,
"train_samples_per_second": 18.218,
"train_steps_per_second": 1.142
}