Key Specifications
| Vendor | alibaba |
|---|
| Version | 2.5-7b |
|---|
| Release Date | 2024-09-19 |
|---|
| Context Window | 131072 tokens |
|---|
| Input Modalities | text |
|---|
| Output Modalities | text |
|---|
| License | Qwen License |
|---|
| Documentation | https://qwenlm.github.io/ |
|---|
Benchmark Performance
| Benchmark | Score | Unit | Evaluated At | Notes | Source |
|---|
| MMLU | 73.6 | % | 2024-09-19 | 5-shot | view |
| HUMANEVAL | 66.5 | pass@1 | 2024-09-19 | — | view |
| GSM8K | 63.7 | % | 2024-09-19 | 0-shot CoT | view |
| MATH | 41.9 | % | 2024-09-19 | 0-shot CoT | view |
| BBH | 61.9 | % | 2024-09-19 | 3-shot CoT | view |
| GPQA | 28.5 | % | 2024-09-19 | 0-shot | view |
| IFEVAL | 55.4 | % | 2024-09-19 | prompt_strict | view |
| ARC | 88.4 | % | 2024-09-19 | challenge | view |
| MUSR | 46.9 | % | 2024-09-19 | 0-shot | view |
| WINOGRANDE | 71.4 | % | 2024-09-19 | 0-shot | view |
Compliance
- Data Residency: CN
- SOC2: ✗
- HIPAA: ✗
- GDPR: ✗
- ISO 27001: ✗
Qwen2.5 7B
Descripción del modelo
阿里巴巴 Qwen2.5 7B 经济型开源模型, 131K 上下文, 7B 参数, 性能超越 Llama 3 8B, 适合本地部署。
Especificaciones principales
| Proveedor | Versión | Fecha de lanzamiento | Ventana de contexto | Modalidades de entrada | Modalidades de salida | Licencia |
|---|
| Alibaba | 2.5-7b | 2024-09-19 | 131K | text | text | Qwen License |
Rendimiento en benchmarks
| Benchmark | Puntuación | Unidad | Notas |
|---|
| MMLU (Massive Multitask Language Understanding) | 73.6 | % | 5-shot |
| HumanEval | 66.5 | pass@1 | — |
| GSM8K (Grade School Math 8K) | 63.7 | % | 0-shot CoT |
| MATH | 41.9 | % | 0-shot CoT |
| BBH (BIG-Bench Hard) | 61.9 | % | 3-shot CoT |
| GPQA | 28.5 | % | 0-shot |
| IFEval | 55.4 | % | prompt_strict |
| ARC | 88.4 | % | challenge |
| MUSR | 46.9 | % | 0-shot |
| WinoGrande | 71.4 | % | 0-shot |
Precios
| Entrada | Salida | Lectura caché | Escritura caché |
|---|
| — | — | — | — |
por millón de tokens
Fortalezas
Debilidades
Casos de uso
Referencias