Code Review Recommendation: Qwen2.5 72B
Code Review → Qwen2.5 72B
Overzicht gebruiksscenario
Qwen2.5 72B by Alibaba is a recommended option for the Code Review use case.
Aanbevolen modellen
| Rang | Leverancier | Contextvenster | Score |
|---|
| #7 | Alibaba | 131K | 37/100 |
Prestatieanalyse
Benchmarkprestaties
| Benchmark | Score |
|---|
| MMLU | 86.1 |
| HUMANEVAL | 86.6 |
| GSM8K | 88.4 |
| MATH | 83.1 |
| BBH | 82.4 |
Sterktes
- MMLU score 86.1, strong knowledge reasoning.
- HumanEval 86.6, excellent code generation.
- GSM8K 88.4, robust math reasoning.
Vereisten
- 需要 alibaba 的 API 密钥
- 输入长度须在 131K 上下文窗口内