two_agent_rdpo_iter_2 / all_results.json
YYYYYYibo's picture
Model save
8097c02 verified
{
"epoch": 0.99,
"train_loss": 0.6840069897813734,
"train_runtime": 41065.6381,
"train_samples": 20000,
"train_samples_per_second": 0.487,
"train_steps_per_second": 0.004
}