mistral10pGrad_1 / all_results.json
terry69's picture
Model save
97e1afa verified
{
"epoch": 0.9990766389658357,
"total_flos": 2.226930684415443e+18,
"train_loss": 0.7396127296243269,
"train_runtime": 26304.3051,
"train_samples": 103932,
"train_samples_per_second": 3.951,
"train_steps_per_second": 0.021
}