zephyr-7b-gpo-iter2 / train_results.json
lole25's picture
Model save
6c1d38b verified
{
"epoch": 2.0,
"train_loss": 0.014390302914878392,
"train_runtime": 11750.1681,
"train_samples": 61135,
"train_samples_per_second": 0.112,
"train_steps_per_second": 0.056
}