Mistral-Nemo-12B-Instruct-SFT / all_results.json
chchen's picture
End of training
71f2116 verified
raw
history blame contribute delete
362 Bytes
{
"epoch": 2.986666666666667,
"eval_loss": 0.10085082799196243,
"eval_runtime": 4.4417,
"eval_samples_per_second": 22.514,
"eval_steps_per_second": 11.257,
"total_flos": 2.198897954832384e+16,
"train_loss": 0.4042978783448537,
"train_runtime": 473.7973,
"train_samples_per_second": 5.699,
"train_steps_per_second": 0.355
}