Any plans on extending the embeddings model API to support multimodal embedding models such as the one from Vertex AI? [https://cloud.google.com/vertex-ai/generative-ai/docs/embeddings/get-multimodal-embeddings](https://cloud.google.com/vertex-ai/generative-ai/docs/embeddings/get-multimodal-embeddings)