b1_math_top_16_1k / train_results.json
ryanmarten's picture
Upload model
7f4b61c verified
{
"epoch": 6.72,
"total_flos": 8.502394284199117e+16,
"train_loss": 0.39509621262550354,
"train_runtime": 3677.0494,
"train_samples_per_second": 1.904,
"train_steps_per_second": 0.019
}