[Doc]: Update outdated note: LMCache now supports chunked prefill

[KevinCheung2259]

📚 The doc issue

In the file examples/offline_inference/cpu_offload_lmcache.py, line 43 states:

# Note that LMCache is not compatible with chunked prefill for now.

This is now outdated. Both vLLM and LMCache have merged PRs to fully support chunked prefill:

vLLM PR #14505, LMCache PR #392

The current note may mislead users into disabling a working feature.

Suggest a potential alternative/fix

Update the comment to either:

# Note: LMCache supports chunked prefill (see vLLM#14505, LMCache#392).

Or remove it entirely if compatibility is now considered stable/default.

Before submitting a new issue...

• Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.