refactor(vllm): take the HF token as HF_TOKEN in .env

[rickstaa]

Standardizes the user-facing Hugging Face token variable on HF_TOKEN — the canonical modern name (HUGGING_FACE_HUB_TOKEN was deprecated by huggingface_hub in 2023) and the same name the api-proxy example (#41) uses, so both examples read one variable from .env.

The container-side name is unchanged: compose maps HUGGING_FACE_HUB_TOKEN=${HF_TOKEN:-}, since that is the variable the stock vLLM image reads. The token stays optional (:- empty default) — only gated models need it.

[Copilot]

Pull request overview

Standardizes the user-facing Hugging Face token environment variable for the vllm example to HF_TOKEN, while keeping the container-facing variable as HUGGING_FACE_HUB_TOKEN (as expected by the vLLM image). This aligns the vllm example with the token naming used elsewhere (e.g., api-proxy) and keeps the token optional.

Changes:

• Update vllm/compose.yml to map HUGGING_FACE_HUB_TOKEN from ${HF_TOKEN:-}.
• Update vllm/.env.example to document HF_TOKEN instead of HUGGING_FACE_HUB_TOKEN.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
vllm/compose.yml	Maps container env HUGGING_FACE_HUB_TOKEN from the standardized user env HF_TOKEN with an empty default.
vllm/.env.example	Renames the documented token variable to HF_TOKEN to match the new user-facing convention.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.