Files
genv3pair1NoGT_1.5B_cdpo_eb…/all_results.json
ModelHub XC b64d4a9f8e 初始化项目,由ModelHub XC社区提供模型
Model: YuchenLi01/genv3pair1NoGT_1.5B_cdpo_ebs32_lr5e-06_beta0.1_epoch16.0_42
Source: Original Platform
2026-10-23 04:25:18 +08:00

22 lines
737 B
JSON

{
"epoch": 16.0,
"eval_logits/chosen": -3.3399085998535156,
"eval_logits/rejected": -3.46624755859375,
"eval_logps/chosen": -47.42058563232422,
"eval_logps/rejected": -75.21272277832031,
"eval_loss": 2.4941890239715576,
"eval_rewards/accuracies": 0.125,
"eval_rewards/chosen": -1.73505437374115,
"eval_rewards/margins": -1.73505437374115,
"eval_rewards/rejected": 0.0,
"eval_runtime": 33.5322,
"eval_samples": 310,
"eval_samples_per_second": 9.245,
"eval_steps_per_second": 0.298,
"total_flos": 0.0,
"train_loss": 0.0022047922514670387,
"train_runtime": 763.2324,
"train_samples": 5700,
"train_samples_per_second": 119.492,
"train_steps_per_second": 3.752
}