Bug Report

Description

Bug Summary:
The conversation recording functionality in chat mode captures the LLM's responses as part of the user's input, resulting in repetition of the LLM's output.

Steps to Reproduce:

1. Start a call
   with the LLM using speakers and a mic (i.e., not headphones)
2. Observe that the LLM's responses are being recorded as part of the conversation from the user's end.

Expected Behavior:
In chat mode, the conversation recording should only capture the user's input and not include the LLM's responses. The LLM's output should be displayed separately and not be treated as part of the user's input.

Actual Behavior:
In chat mode, the conversation recording captures both the user's input and the LLM's responses. This results in the LLM's output being included in the conversation as if it were coming from the user's end, leading to repetition of the LLM's responses.

Environment

• Open WebUI Version: v0.3.7

• Ollama (if applicable): I'm running this with aphrodite

• Operating System: Ubuntu 22.04

• Browser (if applicable): Firefox 127.0.2 (64-bit)

Reproduction Details

Confirmation:

• [x ] I have read and followed all the instructions provided in the README.md.
• [x ] I am on the latest version of both Open WebUI and Ollama.
• I have included the browser console logs.
• I have included the Docker container logs.

Logs and Screenshots

Browser Console Logs:
[Include relevant browser console logs, if applicable]

Docker Container Logs:
[Include relevant Docker container logs, if applicable]

Screenshots (if applicable):

Installation Method

Docker

Additional Information

I've observed this issue consistently across multiple conversations with the LLM. It obviously works fine with headphones. I've tried restarting the chat interface and reloading the page, but the behavior remains the same. Any guidance on how to address this issue would be appreciated.