Independent measured model card
HuggingFaceTB/SmolLM2-360M-Instruct
Hugging Face source ↗ · JSON for agents
361.8M unique parameters (a10cc1512eab)
Every result links to its execution evidence. Compare scores only within matching protocols; no overall best-model claim. GPU-slot time includes evaluator setup and CPU work.
All current benchmark results
| Benchmark / coverage | Result | Version / protocol | GPU-slot time | Evidence |
|---|---|---|---|---|
| ARC-Challenge subset · 32 items | 28.12% accuracy · max | a10cc1512eabcodesota-sub1b-choice-v1 | 4.3s | Inspect run · Manifest |
| ARC-Easy subset · 32 items | 43.75% accuracy · max | a10cc1512eabcodesota-sub1b-choice-v1 | 4.2s | Inspect run · Manifest |
| ARC-Easy (full test split, raw accuracy) full · 2,376 items | 56.52% accuracy · max | a10cc1512eabcodesota-sub1b-arc-easy-standard-v2 | 26.6s | Inspect run · Manifest |
| BoolQ subset · 32 items | 34.38% accuracy · max | a10cc1512eabcodesota-sub1b-choice-v1 | 7.4s | Inspect run · Manifest |
| HellaSwag subset · 32 items | 46.88% accuracy · max | a10cc1512eabcodesota-sub1b-choice-v1 | 4.2s | Inspect run · Manifest |
| OpenBookQA subset · 32 items | 43.75% accuracy · max | a10cc1512eabcodesota-sub1b-choice-v1 | 4.1s | Inspect run · Manifest |