github-starred/open-webui
2
0
Fork 0
mirror of https://github.com/open-webui/open-webui.git synced 2026-08-30 17:25:30 -05:00
Code Issues 2.4k Packages Projects Releases 129 Wiki Activity

Memory leak when embedding #3293

New Issue
Closed
opened 2025-11-11 15:28:16 -06:00 by GiteaMirror · 2 comments
GiteaMirror commented 2025-11-11 15:28:16 -06:00
Owner
Copy Link
Copy Source

Originally created by @Jehyun97 on GitHub (Jan 16, 2025).

Description

After completing the embedding process, I observed a significant increase in RAM usage, which does not decrease over time. I am using [Ollama/OpenWebUI], and I suspect the issue may be related to how RAM is managed after the embedding process. However, when both Ollama and OpenWebUI are turned off, the RAM usage does not increase, indicating that the issue is related to one of these components. Additionally, I installed OpenWebUI using pip.

Steps to Reproduce

  1. Install OpenWebUI via pip install open-webui.
  2. Run the embedding process on a set of documents using [Ollama/OpenWebUI].
  3. Monitor RAM usage during and after the process.
  4. Notice that RAM usage remains high even after the embedding process is completed.

Expected Behavior

RAM usage should decrease or stabilize once the embedding process is completed and resources are released.

Actual Behavior

RAM usage continues to rise or remains elevated after the embedding process is completed.

Environment

  • Hardware: NVIDIA RTX 4090, 128GB RAM
  • Software: [Specify Ollama/OpenWebUI versions]
  • Operating System: [Your OS version]

Additional Information

  • When both Ollama and OpenWebUI are turned off, RAM usage does not increase, which suggests that the issue is specific to one of these components.
  • Is this issue related to caching, session management, or intermediate data not being released?
  • Are there any known methods to free RAM after the embedding process in [Ollama/OpenWebUI]?

Any insights or fixes would be greatly appreciated. Thank you!

Originally created by @Jehyun97 on GitHub (Jan 16, 2025). ### Description After completing the embedding process, I observed a significant increase in RAM usage, which does not decrease over time. I am using [Ollama/OpenWebUI], and I suspect the issue may be related to how RAM is managed after the embedding process. However, when both Ollama and OpenWebUI are turned off, the RAM usage does not increase, indicating that the issue is related to one of these components. Additionally, I installed OpenWebUI using `pip`. ### Steps to Reproduce 1. Install OpenWebUI via `pip install open-webui`. 2. Run the embedding process on a set of documents using [Ollama/OpenWebUI]. 3. Monitor RAM usage during and after the process. 4. Notice that RAM usage remains high even after the embedding process is completed. ### Expected Behavior RAM usage should decrease or stabilize once the embedding process is completed and resources are released. ### Actual Behavior RAM usage continues to rise or remains elevated after the embedding process is completed. ### Environment - Hardware: NVIDIA RTX 4090, 128GB RAM - Software: [Specify Ollama/OpenWebUI versions] - Operating System: [Your OS version] ### Additional Information - When both Ollama and OpenWebUI are turned off, RAM usage does not increase, which suggests that the issue is specific to one of these components. - Is this issue related to caching, session management, or intermediate data not being released? - Are there any known methods to free RAM after the embedding process in [Ollama/OpenWebUI]? Any insights or fixes would be greatly appreciated. Thank you!
GiteaMirror commented 2025-11-11 15:28:19 -06:00
Author
Owner
Copy Link
Copy Source

@tjbck commented on GitHub (Jan 16, 2025):

Are you using the built-in inference engine for embeddings?

@tjbck commented on GitHub (Jan 16, 2025): Are you using the built-in inference engine for embeddings?
GiteaMirror commented 2025-11-11 15:28:19 -06:00
Author
Owner
Copy Link
Copy Source

@Jehyun97 commented on GitHub (Jan 16, 2025):

I am using Ollama as the inference engine for embeddings. The embeddings are generated through the built-in functionality provided by Ollama.

@Jehyun97 commented on GitHub (Jan 16, 2025): I am using Ollama as the inference engine for embeddings. The embeddings are generated through the built-in functionality provided by Ollama.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
bug
confirmed
confirmed issue
core
documentation
enhancement
good first issue
help wanted
non-core
pull-request
Mirrored from GitHub Pull Request
 python
question
testing wanted
No labels
Milestone
No items
No Milestone
Projects
Clear projects
No projects
Assignees
Clear assignees
GiteaMirror ninjasurge (Kaehlen Rog)
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#3293