two_agent_dpo_iter_2 / all_results.json
YYYYYYibo's picture
Model save
a7902e5 verified
{
"epoch": 0.99,
"train_loss": 0.6840333373718013,
"train_runtime": 39783.2449,
"train_samples": 20000,
"train_samples_per_second": 0.503,
"train_steps_per_second": 0.004
}