phi-2-gpo-v25-i1 / train_results.json
lole25's picture
Model save
123afc1 verified
{
"epoch": 2.0,
"train_loss": 0.16023516272887206,
"train_runtime": 6530.1152,
"train_samples": 15000,
"train_samples_per_second": 4.594,
"train_steps_per_second": 0.143
}