Test models locally, starting with a smoke test.
Evaluates Hugging Face Hub models locally and helps choose evaluation frameworks and inference backends.
When to use it
Use for backend selection, local GPU evaluations, and choosing between vLLM, Transformers, and accelerate.
Give it a model ID and evaluation tasks; it runs evaluations and saves results on the local computer.
What you provide
The evaluation scripts require Python 3.10 or newer.
The recommended local execution path uses uv run and checks that uv is available.
This skill
Hugging Face Hub
Loads models from Hugging Face
Hugging Face model repository code
Custom model repository code