| title: Python sandbox math demo | |
| emoji: 🧮 | |
| colorFrom: blue | |
| colorTo: gray | |
| sdk: static | |
| app_file: index.html | |
| pinned: false | |
| license: mit | |
| short_description: Same question with and without a run_python tool. | |
| # Python sandbox math demo | |
| This Space shows a small local model (LFM3-1.7B) answering the same math question twice: once on its own, and once with a sandboxed `run_python` tool. For each side you see the answer, the generated tokens, the time and the tool calls. | |
| **It replays recorded runs. It is not live.** Hugging Face now needs a PRO subscription to host a Gradio Space on free CPU, so this is a static Space. The runs were recorded with `live/app.py` (included here) on a 2-thread Linux machine using the bwrap sandbox backend. | |
| Tool code and results: https://huggingface.co/austindixson/python-sandbox-math-tool | |
| Notes: | |
| - To run it live yourself: `cd live && pip install -r requirements.txt && python app.py`. It downloads llama.cpp b11938 and the LFM3-1.7B Q4_0 GGUF on first start (Linux x64). On 2 CPU cores, expect up to a minute per side. | |
| - At full context (Oct 30 run on an A100), the tool gave about the same accuracy as no tools and used 6 to 23% fewer tokens per solved problem. The accuracy gains in the earlier capped runs were cap effects. Write-up: https://www.destroyrebuild.xyz/blog/lfm3-python-tool-full-context | |
| - The demo caps each answer at 1,024 tokens. That is not the 25,000-token budget used in the recorded evaluation. A small cap exaggerates the gap, because the no-tools side often runs out before it finishes. | |
| - On Linux the sandbox is bubblewrap if user namespaces work, otherwise a weaker fallback. On a Mac the tool uses `sandbox-exec`, which is stronger. | |