@eugr Hi, would it be possible to add recipe support for the following two models?
- Qwen/Qwen3-VL-Embedding-8B (designed for cross-modal information retrieval)
- Qwen/Qwen3-VL-Reranker-8B (for ranking multimodal search results)
I want to setup GraphRAG on a DGX Spark Cluster and use these two models.