the response from o1 is not formatted correctly?

[elvis-cai]

What happened?

we found the response from o1 is not formatted correctly, you can find from the screenshot, not sure if this is related to we are using litellm, will have another test using openai API endpoint directly

Steps to Reproduce

• deploy librechat 0.7.6
• configure to use o1 llm
• ask question

Write a bash script that takes a matrix represented as a string with
format '[1,2],[3,4],[5,6]' and prints the transpose in the same format.

What browsers are you seeing the problem on?

Chrome

Relevant log output

no useful log is printed

Screenshots

Code of Conduct

• I agree to follow this project's Code of Conduct

[elvis-cai]

did another round test to use openai API directly, unfortunately the same issue persists

[danny-avila]

Ask the AI to output the scripts in code blocks.

A lot of those symbols have special formatting in markdown, which is rendered dynamically:

Try copy/pasting the content here on GitHub and you'll see it format the text similarly.

[elvis-cai]

thanks @danny-avila , it works if I ask it to put in code block, one thing to note is in Chatgpt with o1, it formats correctly, not sure if it's related to OpenAI API capability https://chatgpt.com/share/6789e79f-a208-8013-8aad-99a29c99aba6