b1_math_top_1_1k / train_results.json
ryanmarten's picture
Upload model
23d6f78 verified
{
"epoch": 6.72,
"total_flos": 9.868859925541683e+16,
"train_loss": 0.41120058127811976,
"train_runtime": 4376.5107,
"train_samples_per_second": 1.599,
"train_steps_per_second": 0.016
}