two_agent_1_dpo_iter_2 / train_results.json
YYYYYYibo's picture
Model save
87bda1e verified
{
"epoch": 0.99,
"train_loss": 0.6870396254109401,
"train_runtime": 21840.9373,
"train_samples": 20000,
"train_samples_per_second": 0.916,
"train_steps_per_second": 0.007
}