mirror of
https://github.com/open-webui/open-webui.git
synced 2026-03-15 19:40:22 -05:00
Ollama container does not clean memory usage #271
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @tgsca on GitHub (Feb 6, 2024).
Bug Report
Description
Bug Summary:
I am starting both containers, ollama and ollama-webui, with docker compose. This works well but with each interaction it slows down. I analyzed the usage statistics of the containers and recognized that the ollama container does not clean its memory usage. So the memory getting blocked more and more with each interaction.
After restarting the container the memory is cleaned up.
Steps to Reproduce:
Expected Behavior:
After ollama responded to a question, the memory cleans up automatically.
Actual Behavior:
After ollama responded to a question, the memory keeps reserved/used.
Environment
Reproduction Details
Confirmation:
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):
[Attach any relevant screenshots to help illustrate the issue]
Installation Method
docker compose (Docker version 25.0.2, build 29cf629)
Additional Information
[Include any additional details that may help in understanding and reproducing the issue. This could include specific configurations, error messages, or anything else relevant to the bug.]
Note
If the bug report is incomplete or does not follow the provided instructions, it may not be addressed. Please ensure that you have followed the steps outlined in the README.md and troubleshooting.md documents, and provide all necessary information for us to reproduce and address the issue. Thank you!
@justinh-rahb commented on GitHub (Feb 6, 2024):
Out of scope for this project. This is related to caching the model files in memory for some time after loading them. The behaviour is tunable, but again, out of scope for this project, please refer to Ollama.
@tgsca commented on GitHub (Feb 6, 2024):
Thank you for your response. Understand what you mean and was not sure if this could be of interest for this project. But feedback fully understandable. Thx 👍