Is your feature request related to a problem? Please describe.
The chat title seems to be generated withe the same model that was used in the a chat, which can be unnecessarily slow when chatting with large models like llama2:70B, which takes a lot of resources and time just to summarize a chat.
Describe the solution you'd like
Allow to select a separate (e.g. smaller <7B) model to do the summarization for the title (could be set globally) to speed up the title generation. Maybe even autoload/predefine a small model like neural-chat, llama:7B or similar.
Originally created by @malteneuss on GitHub (Dec 11, 2023).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/206
**Is your feature request related to a problem? Please describe.**
The chat title seems to be generated withe the same model that was used in the a chat, which can be unnecessarily slow when chatting with large models like llama2:70B, which takes a lot of resources and time just to summarize a chat.
**Describe the solution you'd like**
Allow to select a separate (e.g. smaller <7B) model to do the summarization for the title (could be set globally) to speed up the title generation. Maybe even autoload/predefine a small model like neural-chat, llama:7B or similar.
Just added with the latest release, please try it out and let me know if you encounter any issues, Thanks!
<!-- gh-comment-id:1874747263 -->
@tjbck commented on GitHub (Jan 3, 2024):
Just added with the latest release, please try it out and let me know if you encounter any issues, Thanks!
<img width="645" alt="image" src="https://github.com/ollama-webui/ollama-webui/assets/25473318/37b3413b-e2e9-4e8c-9ce6-592fea3ff6e7">
Specifying the title auto-generation model seems to default to the current chats model, and ignores (nor saves) my specified model.
<!-- gh-comment-id:1977834387 -->
@Shadowfita commented on GitHub (Mar 5, 2024):
Specifying the title auto-generation model seems to default to the current chats model, and ignores (nor saves) my specified model.
looks like the setting has been moved to /admin/settings under Interface
<!-- gh-comment-id:2861535349 -->
@fivecanal5 commented on GitHub (May 8, 2025):
> Where is this? I could not find it in settings.
looks like the setting has been moved to /admin/settings under Interface
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @malteneuss on GitHub (Dec 11, 2023).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/206
Is your feature request related to a problem? Please describe.
The chat title seems to be generated withe the same model that was used in the a chat, which can be unnecessarily slow when chatting with large models like llama2:70B, which takes a lot of resources and time just to summarize a chat.
Describe the solution you'd like
Allow to select a separate (e.g. smaller <7B) model to do the summarization for the title (could be set globally) to speed up the title generation. Maybe even autoload/predefine a small model like neural-chat, llama:7B or similar.
@tjbck commented on GitHub (Jan 3, 2024):
Just added with the latest release, please try it out and let me know if you encounter any issues, Thanks!
@Shadowfita commented on GitHub (Mar 5, 2024):
Specifying the title auto-generation model seems to default to the current chats model, and ignores (nor saves) my specified model.
@newptcai commented on GitHub (Mar 21, 2025):
Where is this? I could not find it in settings.
@fivecanal5 commented on GitHub (May 8, 2025):
looks like the setting has been moved to /admin/settings under Interface