docs: fix vLLM tensor parallel flag in deploy guide
#24
by
applesilicon
- opened
The vLLM CLI uses --tensor-parallel-size (or -tp) for tensor parallelism, not --tp. The current example in deploy_guidance.md fails with unrecognized arguments: --tp. This PR updates only the vLLM command example; the SGLang example remains unchanged.
Thanks for your fix!
courage17340
changed pull request status to
merged