llama20pGrad_2 / train_results.json
terry69's picture
Model save
cda8996 verified
{
"epoch": 1.0,
"total_flos": 2.1301458550160097e+18,
"train_loss": 0.26981097944264343,
"train_runtime": 10382.4605,
"train_samples": 103932,
"train_samples_per_second": 10.01,
"train_steps_per_second": 0.078
}