Knowledge
BoolQ Leaderboard
BoolQ is a reading-comprehension benchmark of naturally occurring yes/no questions, each paired with a short passage that contains the answer. It tests whether a model can extract and reason over information from real text rather than just recall facts.
Source: epoch33 open models ranked+44 proprietaryData through Aug 2024
Open models ranked on BoolQ
# shows rank among open models / rank overall (including proprietary).
| # | Model | Score |
|---|---|---|
| 1 / 5 | StableBeluga2 · 70B | 89.4% |
| 2 / 6 | Falcon 180B · 180B | 89.0% |
| 3 / 9 | Llama 2 70B HF · 69.0B | 88.6% |
| 4 / 14 | Internlm 20B · 20B | 87.5% |
| 5 / 15 | Mistral 7B v0.1 · 7B | 87.4% |
| 6 / 18 | Qwen 14B · 14.2B | 86.2% |
| 7 / 21 | Gemma 2 9B · 9.2B | 85.7% |
| 8 / 26 | Phi 3.5 MoE Instruct · 41.9B | 84.6% |
| 9 / 30 | Gemma 7B · 8.5B | 83.2% |
| 10 / 31 | Mistral 7B Instruct v0.2 · 7.2B | 83.2% |
| 11 / 32 | Falcon 40B · 41.8B | 83.1% |
| 12 / 33 | Llama 3.1 8B Instruct · 8.0B | 82.8% |
| 13 / 35 | Llama 2 13B HF · 13.0B | 82.4% |
| 14 / 37 | Vicuna 13B V1.3 · 13B | 80.8% |
| 15 / 40 | Chatglm2 6B · 6B | 79.0% |
| 16 / 43 | Phi 3.5 Mini Instruct · 3.8B | 78.0% |
| 17 / 44 | Llama 2 7B HF · 6.7B | 77.9% |
| 18 / 46 | Llama 7B · 6.7B | 76.5% |
| 19 / 47 | Qwen 7B · 7.7B | 76.4% |
| 20 / 50 | Phi 1 5 · 1.4B | 75.8% |
| 21 / 51 | Falcon 7B · 7.2B | 75.3% |
| 22 / 53 | Xgen 7B 8k Base · 7B | 74.3% |
| 23 / 56 | Bloom · 176.2B | 70.4% |
| 24 / 57 | Gemma 2B · 2.5B | 69.4% |
| 25 / 59 | Qwen 1 8B · 1.8B | 68.0% |
| 26 / 60 | Baichuan2 13B Base · 13B | 67.0% |
| 27 / 62 | GPT J 6B · 6B | 65.4% |
| 28 / 64 | GPT Neox 20B · 20.7B | 64.9% |
| 29 / 65 | Internlm 7B · 7B | 64.1% |
| 30 / 66 | Baichuan2 7B Base · 7B | 63.2% |
| 31 / 69 | Gpt2 Xl · 1.6B | 61.8% |
| 32 / 70 | Cerebras GPT 13B · 13B | 61.1% |
| 33 / 72 | Stablelm Tuned Alpha 7B · 7B | 59.0% |
Score vs model size
Which models give the most quality for their size — the ones worth running locally.
- Phi 1 5, 1B, score 75.8% — on the efficiency frontier (best score at its size or smaller).
- Phi 3.5 Mini Instruct, 4B, score 78.0% — on the efficiency frontier (best score at its size or smaller).
- Chatglm2 6B, 6B, score 79.0% — on the efficiency frontier (best score at its size or smaller).
- Mistral 7B v0.1, 7B, score 87.4% — on the efficiency frontier (best score at its size or smaller).
- Internlm 20B, 20B, score 87.5% — on the efficiency frontier (best score at its size or smaller).
- Llama 2 70B HF, 69B, score 88.6% — on the efficiency frontier (best score at its size or smaller).
- StableBeluga2, 70B, score 89.4% — on the efficiency frontier (best score at its size or smaller).
BoolQ: frequently asked questions
- What is the best open LLM on BoolQ?
- StableBeluga2 is the top open model on BoolQ, scoring 89.4%. Among all models tested — including proprietary ones — it ranks #5. The top model overall is T5 11B (Google) at 91.2%.
- What's the best BoolQ model you can run on a 24 GB GPU?
- Internlm 20B is the highest-scoring open model that fits in 24 GB at 4-bit quantization (about 11 GB), scoring 87.5% on BoolQ.
- What's the best BoolQ model you can run on a 12 GB GPU?
- Internlm 20B is the highest-scoring open model that fits in 12 GB at 4-bit quantization (about 11 GB), scoring 87.5% on BoolQ.
- Can open models match proprietary models on BoolQ?
- Not quite on BoolQ: the strongest proprietary model (T5 11B) scores 91.2%, ahead of the best open model (StableBeluga2) at 89.4% — but you can run the open one yourself.
Scores aggregated from epoch. llmrun does not run this benchmark — see the source for methodology, or the about benchmarks for what it measures.