search

opensearch-project / ml-commons

ml-commons provides a set of common machine learning algorithms, e.g. k-means, or linear regression, to help developers build ML related features within OpenSearch.
Apache License 2.0
84 stars 119 forks source link

Add AI connector blueprints for vertexAI text embedding models #2175

Open aman-tandon-30 opened 4 months ago

aman-tandon-30 commented 4 months ago

Is your feature request related to a problem? Add AI connector blueprint for vertexAI text embedding models in addition to existing ones. If it is already a plan in future release, please close this issue.

What solution would you like? A markdown file of AI connector blueprint for vertexAI gecko text embedding models (textembedding-gecko@003 & google/multimodalembedding-001)

What alternatives have you considered? I am trying to write it myself according to OpenSearch Doc here, and also using existing blueprints here

Do you have any additional context? I am able to add the VertexAI embedding connector on my Opensearch 2.11, I would love to help the community, to doing so it will be helful for people to go through the documentation and implement it quickly. I will be quickly able to raise the PR if this is not planned already.

ylwu-amzn commented 4 months ago

Thanks a lot. Feel free to raise PR for that blueprint

aman-tandon-30 commented 3 months ago

@ylwu-amzn Created the PR for Vertex AI embedding endpoint https://github.com/opensearch-project/ml-commons/pull/2268.

I am not able to make the multimodal embedding works, however embedding worked fine, should I update this issue to strike out the google/multimodalembedding-001