mistral_llama_2_code_math_5_full / train_results.json
CharlesLi's picture
Model save
36a9249 verified
raw
history blame contribute delete
243 Bytes
{
"epoch": 0.995475113122172,
"total_flos": 11463536148480.0,
"train_loss": 0.5680296632376585,
"train_runtime": 618.7593,
"train_samples": 15980,
"train_samples_per_second": 5.703,
"train_steps_per_second": 0.178
}