Files
b2_math_length_3k/train_results.json

8 lines
222 B
JSON
Raw Normal View History

{
"epoch": 6.850632911392405,
"total_flos": 9.289277301231452e+17,
"train_loss": 0.3896952799654433,
"train_runtime": 26951.6631,
"train_samples_per_second": 0.821,
"train_steps_per_second": 0.008
}