phi-2-gpo-test-longest-iter-3 / train_results.json
BraylonDash's picture
Model save
5de8f9d verified
{
"epoch": 1.98,
"train_loss": 0.010348274312432735,
"train_runtime": 1408.9952,
"train_samples": 61135,
"train_samples_per_second": 1.419,
"train_steps_per_second": 0.088
}