Jason Weill

After discussing this with my colleagues, a new command, perhaps called `/reset`, would be best for this use case. Thank you for opening this issue!

@michaelchia The idea of having multiple conversations, perhaps in "threads" (Slack or Discord style) or in "conversations" (ChatGPT style) has come up before, but I didn't see an issue articulating...

@Kevin-McIsaac Thank you for opening this issue! What is the raw output of the LLM you used? Is it producing an inline image, for example, or is it linking to...

@mlejva Thanks for reaching out! Would love to see a pull request or other demo showing how Jupyter AI could work with your project.

@startakovsky Thanks for opening this issue! The `-f math` option presumes that the response is entirely in LaTeX format. Descriptive text, as you've seen, is not rendered properly in LaTeX...

Thanks for clarifying! Prompt templates are now customizable, now that #581 has been merged in, so if you find a template for `math` mode that yields better-quality mixed math and...

Let's leave this open; it may be possible to solve this with prompt engineering related to the prompt template, and we would consider pull requests related to this. Thanks again...

@erkin98 Thanks for opening this issue! Have you set up your credentials as described in the Boto3 documentation here? https://boto3.amazonaws.com/v1/documentation/api/latest/guide/credentials.html

#638 suggests making `/generate` available in magic commands as well.

@sundaraa-deshaw Sorry for the delay in responding to this! Typically, we expect language models to respond with code enclosed in triple backticks (```) on lines by themselves, perhaps with a...