Title Generation not working when task model (external) is set as "Current model" and the model have a custom name #2149

Closed
opened 2025-11-11 15:01:12 -06:00 by GiteaMirror · 0 comments
Owner

Originally created by @Simi5599 on GitHub (Sep 21, 2024).

Bug Report

Installation Method

Docker compose

Environment

  • Open WebUI Version: 0.3.23

  • Ollama (if applicable): Not used

  • Operating System: Windows 11, Android (issue not related with operating system)

  • Browser (if applicable): Firefox, Chrome (Android)

Confirmation:

  • I have read and followed all the instructions provided in the README.md.
  • 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.
  • I have provided the exact steps to reproduce the bug in the "Steps to Reproduce" section below.

Expected Behavior:

When you start a new chat with an external model a title for the chat should be automatically generated

Actual Behavior:

The title is not generated and stays on "New chat"

Description

Bug Summary:
When you start a new chat with an external model, a title for the chat should be automatically generated.
This does not happens if i am using an external model (gpt 4o-mini) and the model have been customized (The name has been changed in the workspace section).
If you select a different model then the title will be correctly generated

Reproduction Details

Steps to Reproduce:

  1. Be sure to use external models (For example gpt ones)
  2. Select an external model and change its name (For example gpt4o-mini to Casper)
  3. Go to the Admin panel -> Settings -> Interface and make sure that "External Models" is set to "Current Model"
  4. Chat wth the LMM
  5. The title will not be generated

Logs and Screenshots

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

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

Screenshots/Screen Recordings (if applicable):
Sent a screen recording

https://github.com/user-attachments/assets/432367c2-85c0-4140-b28d-1b57611e2c99

Additional Information

The model i am using is gpt-4o-mini and we have changed the followings:

  • Name
  • System Prompt
  • Temperature = 0.8
  • Context Length = 4096
  • Batch Size = 512
  • num_keep = 24
  • Max Tokens = 4096
  • Some tools are enabled (Run code)
  • Vision = on

EDIT: Also the web search seems to be affected (if enabled): Open-webui asks the model to search my entire prompt

Originally created by @Simi5599 on GitHub (Sep 21, 2024). # Bug Report ## Installation Method Docker compose ## Environment - **Open WebUI Version:** 0.3.23 - **Ollama (if applicable):** Not used - **Operating System:** Windows 11, Android (issue not related with operating system) - **Browser (if applicable):** Firefox, Chrome (Android) **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. - [x] I have provided the exact steps to reproduce the bug in the "Steps to Reproduce" section below. ## Expected Behavior: When you start a new chat with an external model a title for the chat should be automatically generated ## Actual Behavior: The title is not generated and stays on "New chat" ## Description **Bug Summary:** When you start a new chat with an external model, a title for the chat should be automatically generated. This does not happens if i am using an external model (gpt 4o-mini) and the model have been customized (The name has been changed in the workspace section). If you select a different model then the title will be correctly generated ## Reproduction Details **Steps to Reproduce:** 1. Be sure to use external models (For example gpt ones) 2. Select an external model and change its name (For example gpt4o-mini to Casper) 3. Go to the Admin panel -> Settings -> Interface and make sure that "External Models" is set to "Current Model" 4. Chat wth the LMM 5. The title will not be generated ## Logs and Screenshots **Browser Console Logs:** [Include relevant browser console logs, if applicable] **Docker Container Logs:** [Include relevant Docker container logs, if applicable] **Screenshots/Screen Recordings (if applicable):** Sent a screen recording https://github.com/user-attachments/assets/432367c2-85c0-4140-b28d-1b57611e2c99 ## Additional Information The model i am using is gpt-4o-mini and we have changed the followings: - Name - System Prompt - Temperature = 0.8 - Context Length = 4096 - Batch Size = 512 - num_keep = 24 - Max Tokens = 4096 - Some tools are enabled (Run code) - Vision = on EDIT: Also the web search seems to be affected (if enabled): Open-webui asks the model to search my entire prompt
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#2149