opencodereasoning_300k / train_results.json
sedrickkeh's picture
End of training
a7362a7 verified
{
"epoch": 4.994219653179191,
"total_flos": 4.168742448987636e+19,
"train_loss": 0.5089499812619186,
"train_runtime": 355375.6593,
"train_samples_per_second": 2.337,
"train_steps_per_second": 0.005
}