Qwen2 0.5B Instruct vs SmolLM2 135M Instruct
Lado a lado. Cada métrica. Cada benchmark.
| Tipo | Qwen2 0.5B Instruct | SmolLM2 135M Instruct |
|---|---|---|
| Provider |  Alibaba | HF Hugging Face TB |
| puntuación promedio | 6.6 | 15.0 |
| Precio de entrada | - | - |
| Precio de salida | - | - |
| Ventana de contexto | - | - |
| Publicado el | 2024-06-03 | 2024-10-31 |
| Código abierto | Open Source | Open Source |
Puntuaciones de benchmark
6 benchmarks · Qwen2 0.5B Instruct: 3, SmolLM2 135M Instruct: 2
| Benchmark | Categoría | Qwen2 0.5B Instruct | SmolLM2 135M Instruct |
|---|---|---|---|
| BBH (HuggingFace) | general | 5.9 | 4.7 |
| GPQA | knowledge | 0.0 | 0.0 |
| IFEval | language | 22.5 | 28.8 |
| MATH Level 5 | math | 2.9 | 0.3 |
| MMLU-PRO | knowledge | 5.9 | 1.3 |
| MUSR | reasoning | 2.4 | 3.7 |