MMR-DAPO / all_results.json
kangdawei's picture
Model save
d1ab61c verified
{
"total_flos": 0.0,
"train_loss": 0.0197520790877752,
"train_runtime": 32053.2934,
"train_samples": 7000,
"train_samples_per_second": 0.3,
"train_steps_per_second": 0.006
}