Text Generation on FuseEval Chinese Scenario
32.15ROUGE-1SpecEM
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| SpecEMModel Category=Methods of Ensembling Base LLMs2024.12 | 32.15 | 9.94 | 24 | 4.75 | -4.29 | 71.03 | 3.25 | |
| Qwen2-72B-instructModel Category=Larger LLMs2024.12 | 31.44 | 8.97 | 22.48 | 4.88 | -4.34 | 70.65 | 3.63 | |
| Mistral-7B-instruct-v0.3Model Category=Base LLMs2024.12 | 30.99 | 8.65 | 20.66 | 4.42 | -4.48 | 70.1 | 6.55 | |
| MOAModel Category=Methods of Ensembling Base LLMs2024.12 | 30.96 | 8.5 | 20.6 | 4.36 | -4.31 | 70.11 | 5.93 | |
| MBRModel Category=Methods of Ensembling Base LLMs2024.12 | 30.93 | 8.71 | 20.63 | 4.31 | -4.31 | 70.23 | 5.33 | |
| Glm-4-9B-instructModel Category=Base LLMs2024.12 | 30.88 | 8.71 | 20.42 | 4.47 | -4.3 | 70.25 | 5.18 | |
| GF (Mistral)Model Category=Methods of Ensembling Base LLMs2024.12 | 30.29 | 8.12 | 20.33 | 3.88 | -4.54 | 70.04 | 7.28 | |
| GF (Glm-4)Model Category=Methods of Ensembling Base LLMs2024.12 | 30.26 | 8.7 | 20.51 | 4.27 | -4.33 | 70.23 | 5.4 | |
| Qwen2-7B-instructModel Category=Base LLMs2024.12 | 29.93 | 8.09 | 20.03 | 3.62 | -4.33 | 69.99 | 6.42 | |
| Gemma-2-9B-instructModel Category=Base LLMs2024.12 | 29.15 | 7.65 | 18.35 | 3.36 | -4.28 | 68.73 | 8.58 | |
| GF (Qwen2)Model Category=Methods of Ensembling Base LLMs2024.12 | 28.69 | 7.87 | 18.93 | 3.32 | -4.41 | 69.81 | 8.41 | |
| Llama-3-70B-instructModel Category=Larger LLMs2024.12 | 27.78 | 7.05 | 20.22 | 4.14 | -4.55 | 68.52 | 7.38 | |
| UniTEModel Category=Methods of Ensembling Base LLMs2024.12 | 27.46 | 8.22 | 20.54 | 2.83 | -4.6 | 67.58 | 9.37 |