Integration Support for GPT-4 Turbo and Updated GPT 3.5 Turbo Models

[alex-feel]

I'd like to request the integration of the new GPT-4 Turbo (gpt-4-1106-preview) and Updated GPT 3.5 Turbo (gpt-3.5-turbo-1106) models as announced in OpenAI's recent blog post.

The current documentation for litellm mentions support for only gpt-4-1106-preview. It would be beneficial for us to support both models mentioned above.

Please note, this might require an update of the openai module to incorporate bug fixes from this PR. However, the latest litellm package version 0.13.2 is compatible with up to openai package version 0.28.1, which should be taken into account to ensure compatibility is maintained.

[yuhongsun96]

Good suggestion, will look into it and also ask the LiteLLM folks

[photonn]

In Azure, gpt-4-1106-preview IS gpt-4 turbo. I'm using it with danswer without issues since a few weeks ago:

https://techcommunity.microsoft.com/t5/ai-azure-ai-services-blog/azure-openai-service-launches-gpt-4-turbo-and-gpt-3-5-turbo-1106/ba-p/3985962

Is curious as Microsoft announced 120k token window in ignite but I have only 80k available. Maybe they make a mistake with the window of 3.5-turbo, which is 120k

[yuhongsun96]

Yup, gpt-4-turbo has been supported for quite some time now, going to close this one now!