Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

adding support for cross region models #246

Closed
wants to merge 1 commit into from

Conversation

rsgrewal-aws
Copy link

This PR is rebased from the approved PR #187

this includes the fixes and is re-based to the main branch

@rsgrewal-aws
Copy link
Author

@3coins could you please review this PR -- this PR has been taken from #187 which has been approved and rebased to the main branch. we have a few customers asking for this

@3coins
Copy link
Collaborator

3coins commented Oct 21, 2024

@rsgrewal-aws
#242 takes care of the cross-region inference ids without hard-coding the region prefixes as recommended by the Bedrock team. I am doing some validation integration checks on that PR, will be releasing this within a day or two.

@3coins 3coins closed this Oct 21, 2024
@rsgrewal-aws
Copy link
Author

@3coins why do we need the provider values and why do we need to parse the model id ? can we simply use the model id as is ?

@3coins
Copy link
Collaborator

3coins commented Oct 21, 2024

@rsgrewal-aws
We use the model-id as is, but need the provider to determine the value to add for the disable_streaming attribute.

@rsgrewal-aws
Copy link
Author

agreed but we try to parse the model id to get the streaming value -- for converse api we have to assume streaming is supported and for ChatBedrock we can have model_args being passed in -- one of the keys in this can be provider, second can be streaming . this will avoid any future issues of model id/arn etc

@3coins
Copy link
Collaborator

3coins commented Oct 22, 2024

@rsgrewal-aws

for converse api we have to assume streaming is supported

As per the documentation, only Anthropic and Cohere support streaming with tools at the moment.
https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference-supported-models-features.html

one of the keys in this can be provider, second can be streaming

What do you mean by this?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants