MMLU College Medicine (Open Medical-LLM)
Science · 2024-04-19
The college-medicine subset of MMLU, 173 questions, scored by the Open Medical-LLM Leaderboard under lm-evaluation-harness with a fixed prompt.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-4 | 76.9 |
| Llama 3 8B Instruct | 65.3 |
| Solar 10.7B Instruct V1.0 | 65.3 |
| GPT-3.5 Turbo 1106 | 64.7 |
| Llama 3 8B | 61.8 |
| Qwen1.5 7B | 60.7 |
| Gemma 7B | 60.7 |
| Zephyr 7B Beta | 60.7 |
| Mistral 7B v0.1 Base | 59.5 |
| Qwen1.5 7B Chat | 58.4 |
| Gemma 1.1 7B IT | 56.1 |
| Mistral 7B Instruct v0.1 | 53.8 |
| Gemma 7B IT | 42.8 |
| Phi-1.5 | 37.6 |
| Gemma 2B | 31.2 |
| Falcon 7B | 24.9 |