jackf857/llama-3-8b-base-cpo-ultrafeedback-4xH200-batch-128-rerun
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: jackf857/llama-3-8b-base-cpo-ultrafeedback-4xH200-batch-128-rerun
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-05-09 21:05:35 +08:00
commit 788b05c009
19 changed files with 3454 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
eval_results.json Normal file
View File

@@ -0,0 +1,17 @@
{
"epoch": 0.9989528795811519,
"eval_logits/chosen": -0.7155267000198364,
"eval_logits/rejected": -0.7125465869903564,
"eval_logps/chosen": -272.6193542480469,
"eval_logps/rejected": -267.1548767089844,
"eval_loss": 2.030885934829712,
"eval_nll_loss": 0.9492784738540649,
"eval_rewards/accuracies": 0.5180000066757202,
"eval_rewards/chosen": -2.7261929512023926,
"eval_rewards/margins": -0.05464465916156769,
"eval_rewards/rejected": -2.67154860496521,
"eval_runtime": 40.4909,
"eval_samples": 2000,
"eval_samples_per_second": 49.394,
"eval_steps_per_second": 3.087
}