Hugging Face's logo

Reward-Reasoning
/
RRM-7B

Safetensors
Model card Files
xet
Community
1
Reward-Reasoning commited on
Commit
8a900ee
·
verified ·
1 Parent(s): ac5542f

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -83,7 +83,7 @@ RRMs have been extensively evaluated on several benchmarks.
83
 
84
  | Model | MMLU-Pro | MATH | GPQA | Overall |
85
  |----------------------------|----------|---------|---------|---------|
86
- | Skywork-Reward-Gemma-2-27B-v0.2 | 55.0 | 46.2 | 44.7 | 48.6 |
87
  | J1-Llama-8B (SC@32) | 67.5 | 76.6 | 55.7 | 66.7 |
88
  | J1-Llama-70B (SC@32) | 79.9 | 88.1 | 66.5 | 78.2 |
89
  | DeepSeek-GRM-27B (MetaRM) (voting@32) | 68.1 | 70.0 | 56.9 | 65.0 |
 
83
 
84
  | Model | MMLU-Pro | MATH | GPQA | Overall |
85
  |----------------------------|----------|---------|---------|---------|
86
+ | Skywork-Reward-Gemma-2-27B | 55.0 | 46.2 | 44.7 | 48.6 |
87
  | J1-Llama-8B (SC@32) | 67.5 | 76.6 | 55.7 | 66.7 |
88
  | J1-Llama-70B (SC@32) | 79.9 | 88.1 | 66.5 | 78.2 |
89
  | DeepSeek-GRM-27B (MetaRM) (voting@32) | 68.1 | 70.0 | 56.9 | 65.0 |