Reading Comprehension on RACE-h (Score)
38.56RACE-h ScoreRWKV-7B
Evaluation results
| Method | Links | |
|---|---|---|
| RWKV-7BBackbone=RWKV-7B, Pruning ratio=0%2024.03 | 38.56 | |
| LLM-StreamlineTrain-Free=false, Base Model=Llama 2 7B, Compression Ratio=25%2025.05 | 37 | |
| ReplaceMe (Cosine)Train-Free=true, Base Model=Llama 2 7B, Compression Ratio=25%2025.05 | 36.7 | |
| ReplaceMe (LS)Train-Free=true, Base Model=Llama 2 7B, Compression Ratio=25%2025.05 | 36.5 | |
| Llama 2 7B (baseline)Base Model=Llama 2 7B, Compression Ratio=0%2025.05 | 35.5 | |
| UIDLTrain-Free=false, Base Model=Llama 2 7B, Compression Ratio=25%2025.05 | 34.7 | |
| SliceGPTTrain-Free=false, Base Model=Llama 2 7B, Compression Ratio=25%2025.05 | 29.4 | |
| ShortGPTBackbone=RWKV-7B, Pruning ratio=9.4%2024.03 | 25.9 | |
| Mamba-2.8BBackbone=Mamba-2.8B, Pruning ratio=0%2024.03 | 24.9 | |
| ShortGPTBackbone=Mamba-2.8B, Pruning ratio=10.9%2024.03 | 24.27 | |
| ShortGPTBackbone=Mamba-2.8B, Pruning ratio=25%2024.03 | 23.67 | |
| ShortGPTBackbone=Mamba-2.8B, Pruning ratio=20.3%2024.03 | 23.61 | |
| LaCoTrain-Free=false, Base Model=Llama 2 7B, Compression Ratio=25%2025.05 | 22.6 | |
| ShortGPTBackbone=RWKV-7B, Pruning ratio=25%2024.03 | 22.56 | |
| ShortGPTBackbone=RWKV-7B, Pruning ratio=28.1%2024.03 | 22.27 | |
| ShortGPTBackbone=RWKV-7B, Pruning ratio=18.8%2024.03 | 22.04 | |
| LLMPrunerTrain-Free=false, Base Model=Llama 2 7B, Compression Ratio=25%2025.05 | 22 | |
| ShortGPTBackbone=Mamba-2.8B, Pruning ratio=31.3%2024.03 | 21.87 |