Sagemaker inference endpoint support

[twerkmeister]

Describe the feature or potential improvement

Currently, when adding llm connections in the UI one can choose between adapters for openai, anthropic, azure, bedrock, google ai vertex, and google ai studio.

We are hosting the finetuned model for our most high volume usecase on sagemaker inference endpoints and as far as I can see we cannot add it as a llm connection to run experiments on datasets comparing different versions of that model.

Is this adapter planned or is it possible to make it work through clever configuration of some other adapter?

Additional information

No response