c1_math_0d_4s_10k / train_results.json
ryanmarten's picture
Upload model
b58d38d verified
{
"epoch": 4.992,
"total_flos": 1.8773457441390592e+18,
"train_loss": 0.3770592579474816,
"train_runtime": 19616.6723,
"train_samples_per_second": 2.549,
"train_steps_per_second": 0.02
}