a1_math_hendrycks / all_results.json
ryanmarten's picture
End of training
0b89df1 verified
raw
history blame
224 Bytes
{
"epoch": 4.988354430379747,
"total_flos": 1.4390160925078323e+18,
"train_loss": 0.21472142088582846,
"train_runtime": 51117.6952,
"train_samples_per_second": 3.091,
"train_steps_per_second": 0.024
}