openr1_codeforces_10k / train_results.json
ryanmarten's picture
End of training
14c026e verified
{
"epoch": 4.992,
"total_flos": 1.891704504856871e+18,
"train_loss": 0.5282762617636949,
"train_runtime": 51076.3739,
"train_samples_per_second": 0.979,
"train_steps_per_second": 0.008
}