Files
tinyllama-1.1b-dpo-pku-safe…/eval_results.json
ModelHub XC edd7af9b86 初始化项目,由ModelHub XC社区提供模型
Model: AIPlans/tinyllama-1.1b-dpo-pku-saferlhf_2
Source: Original Platform
2026-06-11 03:16:17 +08:00

15 lines
562 B
JSON

{
"epoch": 0.9997600191984641,
"eval_logits/chosen": -2.6097218990325928,
"eval_logits/rejected": -2.4986746311187744,
"eval_logps/chosen": -234.91114807128906,
"eval_logps/rejected": -217.71800231933594,
"eval_loss": 0.4486246407032013,
"eval_rewards/accuracies": 0.800000011920929,
"eval_rewards/chosen": -0.17781423032283783,
"eval_rewards/margins": 1.0409064292907715,
"eval_rewards/rejected": -1.218720555305481,
"eval_runtime": 21.1512,
"eval_samples_per_second": 47.279,
"eval_steps_per_second": 11.82
}