litellm/llms/hosted_vllm/embedding/README.md

No transformation is required for hosted_vllm embedding.

VLLM is a superset of OpenAI's embedding endpoint.

To pass provider-specific parameters, see this