TinyQA Benchmark++ on tinyqabenchmark_core-en
Metric: Exact Match (higher is better)
Metric: Exact Match (higher is better)
| # | Model↕ | Exact Match▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | gemma-3-4b | 86.5 | No | Tiny QA Benchmark++: Ultra-Lightweight, Syntheti... | 2025-05-17 | Code |
| 2 | mistral-24b-instruct | 84.6 | No | Tiny QA Benchmark++: Ultra-Lightweight, Syntheti... | 2025-05-17 | Code |
| 3 | llama-3.2-3b-instruct | 84.6 | No | Tiny QA Benchmark++: Ultra-Lightweight, Syntheti... | 2025-05-17 | Code |
| 4 | ministral-8b | 80.8 | No | Tiny QA Benchmark++: Ultra-Lightweight, Syntheti... | 2025-05-17 | Code |
| 5 | ministral-3b | 76.9 | No | Tiny QA Benchmark++: Ultra-Lightweight, Syntheti... | 2025-05-17 | Code |
| 6 | llama-3.2-1b-instruct | 53.8 | No | Tiny QA Benchmark++: Ultra-Lightweight, Syntheti... | 2025-05-17 | Code |
| 7 | mistral-7b-instruct | 50 | No | Tiny QA Benchmark++: Ultra-Lightweight, Syntheti... | 2025-05-17 | Code |