reasoning_0_chat / train_results.json
sedrickkeh's picture
End of training
1cea212 verified
{
"epoch": 4.990403071017274,
"total_flos": 3.115960359367213e+18,
"train_loss": 0.34092234334120386,
"train_runtime": 73926.9587,
"train_samples_per_second": 3.382,
"train_steps_per_second": 0.026
}