feat: completed the vertexai provider integration

[Gmin2]

Completed the integration of vertexAI

Screencast.from.24-12-24.03.01.59.PM.IST.webm

Integration test

Unit test

• Implementation of Bedrock provider
• Unit tests with good coverage
• Integration tests with actual API calls (using test credentials)
• Documentation updates
• Example configuration
• Successful chat completion, completion, and embedding requests
• Error handling and proper status code mapping
• PR review and approval

fix #19
/claim #19

[nirga]

Thanks @Gmin2! Can you add support for streaming as well?

[Gmin2]

Thanks @Gmin2! Can you add support for streaming as well?

Can you review this @nirga, all things done(see the video attached with pr body)

[galkleinman]

Hey @Gmin2,

Please copy, the "Definition of Done" defined in the issue to the PR, and mark everything you've done.

[Gmin2]

Please copy, the "Definition of Done" defined in the issue to the PR, and mark everything you've done.

hey @galkleinman, what it is meant by Example Configuration means do i need to show it in README.md? for docs i need to send PR to docs repo right ?

[galkleinman]

@Gmin2 both in docs and in here

[Gmin2]

can i get a review
cc @galkleinman @nirga

[Gmin2]

Implemented function calling, image, video and audio capabilities of the vertexai provider

cc @galkleinman @nirga

[galkleinman]

Some major changes requests (before I deep dive into the implementation):

• The whole point of this hub/router, is that you use the OpenAI sdk/api to call w/e provider/model you want. The routing is based on a selected pipeline header, which can either support the functionality (either chat, vision, audio...) or not. While you can multiple pipelines support the same functionality with different models or pipeline for team etc... Each pipeline, should expose the same API as openai exposes and implement it internally. With the current implementation you've changed the routing mechanism of the hub and the exposed api.
• Tests aren't reproducible since they count on existence of api keys, models and therefore can't be used in CI. use surf-vcr or any preferred network recording solution to create test recordings which will make it reproducible .

[CLAassistant]

All committers have signed the CLA.

[Gmin2]

The whole point of this hub/router, is that you use the OpenAI sdk/api to call w/e provider/model you want. The routing is based on a selected pipeline header, which can either support the functionality (either chat, vision, audio...) or not. While you can multiple pipelines support the same functionality with different models or pipeline for team etc... Each pipeline, should expose the same API as openai exposes and implement it internally. With the current implementation you've changed the routing mechanism of the hub and the exposed api.

I will do the multimodality part in a seperate PR

• Tests aren't reproducible since they count on existence of api keys, models and therefore can't be used in CI. use surf-vcr or any preferred network recording solution to create test recordings which will make it reproducible .

Done

Can i get a review for now ?

cc @nirga

[nirga]

Hey @Gmin2 can you take a look at the tests? They seem to be failing

[Gmin2]

Hey @Gmin2 can you take a look at the tests? They seem to be failing

Fixed now @nirga