Conversation
|
Converting from draft, utilising the updated client_utils, now. I think this is ready for review. |
Hk669
left a comment
There was a problem hiding this comment.
we can retire this notebook https://microsoft.github.io/autogen/docs/topics/non-openai-models/cloud-togetherai/
Yes, I think you're right. I'll go through and tidy up references to it. Can someone provide guidance on what we do if we remove a document page, e.g. External links may break |
|
…xception capture, warning for no cost, reuse in tests, cleaner tests
|
Thanks @yiranwu0 for working through the code. I've updated most items, just a couple of conversations to discuss and resolve. |
|
@Hk669, I'll wait for feedback on removing the old notebook. I'm thinking perhaps we could just replace the old notebook for this and Mistral with the new ones. What do you think? |
Yeah sounds good. |
All done :)... same in both Mistral and Together PRs |
qingyun-wu
left a comment
There was a problem hiding this comment.
Looks great! Thanks @marklysze!
|
I'm waiting for a few clarifications on the comments made by yiran. Otherwise it looks good to me. |
Co-authored-by: Qingyun Wu <qingyun0327@gmail.com>
Thanks for all your help!! |
|
Resolve the conflict and we are good to go! |
This is a Together AI Client that provides a way to use the Together AI cloud inference service, which provides options to use many LLMs, such as Google's Gemma, Meta's Llama 2/3, Qwen, Mistral.AI's Mistral, and Mixtral, NousResearch's Hermes models.
This client supports text generation and function/tool calling (when using certain models as noted by Together AI).
Token costs are also implemented.
Will be used in the upcoming blog post #2965.
Related issue number
Can be used for #2916.
Checks