Files
llama-3-8b-base-new-dpo-har…/eval_results.json
ModelHub XC e7a907fbc6 初始化项目,由ModelHub XC社区提供模型
Model: jackf857/llama-3-8b-base-new-dpo-harmless-s_star0.6-q_t0.4
Source: Original Platform
2026-06-12 13:30:30 +08:00

17 lines
632 B
JSON

{
"epoch": 0.999244142101285,
"eval_fcm_dpo/beta": 0.00873605813831091,
"eval_logits/chosen": 0.7100939750671387,
"eval_logits/rejected": 0.6652560830116272,
"eval_logps/chosen": -206.1328125,
"eval_logps/ref_chosen": -74.85946655273438,
"eval_logps/ref_rejected": -79.54898834228516,
"eval_logps/rejected": -266.35693359375,
"eval_loss": 0.5597169995307922,
"eval_margin_dpo/margin_mean": 55.534568786621094,
"eval_margin_dpo/margin_std": 95.7474365234375,
"eval_runtime": 38.6273,
"eval_samples": 2303,
"eval_samples_per_second": 59.621,
"eval_steps_per_second": 1.864
}