phi-2-gpo-v8-i1 / train_results.json
lole25's picture
Model save
94c355f verified
{
"epoch": 1.0,
"train_loss": 0.23145956670443216,
"train_runtime": 17376.0848,
"train_samples": 30000,
"train_samples_per_second": 1.727,
"train_steps_per_second": 0.108
}