mistral10pGrad / all_results.json
terry69's picture
Model save
f9eca7f verified
{
"epoch": 1.0,
"total_flos": 2.1311021010303058e+18,
"train_loss": 0.7430525620289037,
"train_runtime": 157132.3963,
"train_samples": 103932,
"train_samples_per_second": 0.661,
"train_steps_per_second": 0.003
}