feat: add VLLMMultivectorDocumentEmbedder and VLLMMultivectorTextEmbedder components - #4014

Hi @maxdswain, thanks for your interest in contributing to Haystack! 🙏

⚠️ You currently have 3 open pull requests in this repository (#3999, #3870 and this one). Our review capacity is limited, so please hold off opening more PRs until we've had a chance to review your first 2 open PRs. This helps us give each contribution the attention it deserves. Thank you!

This is an automated message to help us keep the review queue healthy.

Coverage report (vllm)

Click to see where and how coverage changed

File	Statements	Missing	Coverage	Coverage (new stmts)	Lines missing
integrations/vllm/src/haystack_integrations/components/embedders/vllm
multivector_document_embedder.py
multivector_text_embedder.py
Project Total

This report was generated by python-coverage-comment-action

Hi @anakin87 sorry to bother you, but have you got time to review this? I'm not sure if this has passed you/been missed.

@maxdswain thank you for the ping.

TBH, before implementing components supporting these capabilities, I'd prefer to investigate the overall design of multi-vector retrieval a bit, so I've opened deepset-ai/haystack#13257. Otherwise, I am afraid we will create pieces that don't work well together.

Hi @anakin87 thanks for the response, I'll move this PR into draft as I'd like to contribute more PRs but can only have one open at time.

In terms of multivector support, I was planning on opening PRs for Qdrant and Milvius to update the document writers, and add multivector/late-attention retrievers. It would be nice if multivector was a optional field of Document similar to dense/sparse embeddings.