Now that embedding gemma is released and webllm has added support for gemma 3n. would be great if both are supported for tandem use cases for knowledge and context use cases https://developers.googleblog.com/en/introducing-embeddinggemma/