Vicki Boykis

@msaroufim Feel free to assign this to me and make any edits/clarifications to the text.

I'd like to take this. Just to clarify, we should update all requirements files (starting [with here](https://github.com/pytorch/examples/blob/main/distributed/rpc/pipeline/requirements.txt), to `torch==1.9.0`. I did a quick check of ``` find . -name '*requirements.txt'...

Hey @kimfalk bump on this. Wanted to make sure you saw it and had a chance to reply with any questions/concerns. Thanks!

Great, thanks @kimfalk!

That's awesome, @kimfalk. Excited to try it out. 😄

Hi all, happy to help with this since I'm more familiar with the `OpenaiChatCompletionsLM` class after working on it a bit. Let me know if you'd like support or scaffolding...

Something I'm missing in understanding: If in the task we only use the logprobs for the response for the task evaluation, what does computing the logprobs of the input string...

Thanks, this is super helpful. I also just tested on Together and am getting the same issue, because it's OpenAICompletions compatible: ```python from openai import OpenAI system_content = "You are...

>Further, since OpenAI only gives up to 5 token logprobs I think, we are limited by this—if the logprobs for our next desired target token do not appear in the...

Ah ok so these changes would be to enable local-completions, which is not the same as local-chat-completions, but do have the ability to run echo=True (for legacy reasons), to also...