zephyr-7b-gpo-iter1 / train_results.json
lole25's picture
Model save
5941df8 verified
{
"epoch": 2.0,
"train_loss": 0.005910432943725027,
"train_runtime": 19053.6119,
"train_samples": 61135,
"train_samples_per_second": 0.105,
"train_steps_per_second": 0.052
}