qwen2_7b_best_params / all_results.json
saim1212's picture
second model upload
b072d95 verified
raw
history blame
210 Bytes
{
"epoch": 25.0,
"total_flos": 2.502555032784732e+17,
"train_loss": 0.12197050291108899,
"train_runtime": 13777.8614,
"train_samples_per_second": 0.907,
"train_steps_per_second": 0.091
}