github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #2696] ollama process on macOS using up a lot of RAM while being idle #1609

New Issue
Closed
opened 2026-04-12 11:32:15 -05:00 by GiteaMirror · 11 comments
GiteaMirror commented 2026-04-12 11:32:15 -05:00
Owner
Copy Link

Originally created by @siikdUde on GitHub (Feb 22, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/2696

Originally assigned to: @dhiltgen on GitHub.

SCR-20240222-ozbm

As you can see, ollama is the second most resource intensive application. I am not actively running any models, just the app is open. Any idea why this is?

Originally created by @siikdUde on GitHub (Feb 22, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/2696 Originally assigned to: @dhiltgen on GitHub. <img width="1081" alt="SCR-20240222-ozbm" src="https://github.com/ollama/ollama/assets/10148714/575001a0-9b9a-4e08-ba8c-f0321ec3e6df"> As you can see, ollama is the second most resource intensive application. I am not actively running any models, just the app is open. Any idea why this is?
GiteaMirror added the bugmacos labels 2026-04-12 11:32:15 -05:00
GiteaMirror commented 2026-04-12 11:32:17 -05:00
Author
Owner
Copy Link

@stevengans commented on GitHub (Feb 23, 2024):

If you just launch Ollama it will not take up that memory. However, if you load a model and then close the terminal, the memory will still be used until you close Ollama and relaunch it.

<!-- gh-comment-id:1960802198 --> @stevengans commented on GitHub (Feb 23, 2024): If you just launch Ollama it will not take up that memory. However, if you load a model and then close the terminal, the memory will still be used until you close Ollama and relaunch it.
GiteaMirror commented 2026-04-12 11:32:17 -05:00
Author
Owner
Copy Link

@easp commented on GitHub (Feb 23, 2024):

I'm guessing you probably last ran a fairly large model. Did you give it 5 minutes? I just ran mixtral, when it was done there was a ~1GB ollama process. I came back 5 minutes later and it was gone because it automatically shuts the model down after 5 minutes of inactivity.

Have you seen it using ~1GB+ after longer idle periods?

<!-- gh-comment-id:1962097071 --> @easp commented on GitHub (Feb 23, 2024): I'm guessing you probably last ran a fairly large model. Did you give it 5 minutes? I just ran mixtral, when it was done there was a ~1GB ollama process. I came back 5 minutes later and it was gone because it automatically shuts the model down after 5 minutes of inactivity. Have you seen it using ~1GB+ after longer idle periods?
GiteaMirror commented 2026-04-12 11:32:17 -05:00
Author
Owner
Copy Link

@siikdUde commented on GitHub (Feb 25, 2024):

@stevengans @easp

thanks for the replies. I am now on v0.1.27 and the issue is even worse. I closed out of llava 2 hours ago... I hit Cmd + D and Cmd + C to exit models.

ollama memory
<!-- gh-comment-id:1963103793 --> @siikdUde commented on GitHub (Feb 25, 2024): @stevengans @easp thanks for the replies. I am now on v0.1.27 and the issue is even worse. I closed out of llava 2 hours ago... I hit Cmd + D and Cmd + C to exit models. <img width="1014" alt="ollama memory" src="https://github.com/ollama/ollama/assets/10148714/44d76f79-8f65-474e-83eb-83dc8190bbdf">
GiteaMirror commented 2026-04-12 11:32:18 -05:00
Author
Owner
Copy Link

@oldgithubman commented on GitHub (Mar 30, 2024):

If you just launch Ollama it will not take up that memory. However, if you load a model and then close the terminal, the memory will still be used until you close Ollama and relaunch it.

This is false. Just launching ollama uses a significant amount of VRAM while idle for me. On every card (this is on Linux)

<!-- gh-comment-id:2027907442 --> @oldgithubman commented on GitHub (Mar 30, 2024): > If you just launch Ollama it will not take up that memory. However, if you load a model and then close the terminal, the memory will still be used until you close Ollama and relaunch it. This is false. Just launching ollama uses a significant amount of VRAM while idle for me. On every card (this is on Linux)
GiteaMirror commented 2026-04-12 11:32:18 -05:00
Author
Owner
Copy Link

@easp commented on GitHub (Mar 30, 2024):

@oldmanjk This issue is about MacOS

<!-- gh-comment-id:2028445972 --> @easp commented on GitHub (Mar 30, 2024): @oldmanjk This issue is about MacOS
GiteaMirror commented 2026-04-12 11:32:19 -05:00
Author
Owner
Copy Link

@oldgithubman commented on GitHub (Mar 30, 2024):

@oldmanjk This issue is about MacOS

I'm aware. That's why I mentioned my comment was on Linux. Sometimes issues are cross-platform

<!-- gh-comment-id:2028446520 --> @oldgithubman commented on GitHub (Mar 30, 2024): > @oldmanjk This issue is about MacOS I'm aware. That's why I mentioned my comment was on Linux. Sometimes issues are cross-platform
GiteaMirror commented 2026-04-12 11:32:20 -05:00
Author
Owner
Copy Link

@stevengans commented on GitHub (Mar 30, 2024):

@oldmanjk I would recommend creating a new issue for Linux, as @easp said, these comments are around MacOS.

<!-- gh-comment-id:2028449426 --> @stevengans commented on GitHub (Mar 30, 2024): @oldmanjk I would recommend creating a new issue for Linux, as @easp said, these comments are around MacOS.
GiteaMirror commented 2026-04-12 11:32:21 -05:00
Author
Owner
Copy Link

@oldgithubman commented on GitHub (Mar 30, 2024):

@oldmanjk I would recommend creating a new issue for Linux, as @easp said, these comments are around MacOS.

There are already several issues around this. I recommend you guys read them. You might discover this problem isn't unique to macos

<!-- gh-comment-id:2028449835 --> @oldgithubman commented on GitHub (Mar 30, 2024): > @oldmanjk I would recommend creating a new issue for Linux, as @easp said, these comments are around MacOS. There are already several issues around this. I recommend you guys read them. You might discover this problem isn't unique to macos
GiteaMirror commented 2026-04-12 11:32:22 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (May 2, 2024):

We've changed to a subprocess model in the past few versions which likely resolves this when the model unloads. Are people still seeing a large footprint when idle on 0.1.33?

<!-- gh-comment-id:2091826124 --> @dhiltgen commented on GitHub (May 2, 2024): We've changed to a subprocess model in the past few versions which likely resolves this when the model unloads. Are people still seeing a large footprint when idle on 0.1.33?
GiteaMirror commented 2026-04-12 11:32:23 -05:00
Author
Owner
Copy Link

@oldgithubman commented on GitHub (May 2, 2024):

We've changed to a subprocess model in the past few versions which likely resolves this when the model unloads. Are people still seeing a large footprint when idle on 0.1.33?

The problem seems to have been resolved for me since 0.1.32 (again, in Linux). 0.1.33 isn't released yet, right?

<!-- gh-comment-id:2091873934 --> @oldgithubman commented on GitHub (May 2, 2024): > We've changed to a subprocess model in the past few versions which likely resolves this when the model unloads. Are people still seeing a large footprint when idle on 0.1.33? The problem seems to have been resolved for me since 0.1.32 (again, in Linux). 0.1.33 isn't released yet, right?
GiteaMirror commented 2026-04-12 11:32:23 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (May 5, 2024):

I'm going to close this as resolved. If anyone is still seeing unexpected memory usage behavior when the system is idle and unloads the model, please provide more details and I'll re-open.

<!-- gh-comment-id:2094908805 --> @dhiltgen commented on GitHub (May 5, 2024): I'm going to close this as resolved. If anyone is still seeing unexpected memory usage behavior when the system is idle and unloads the model, please provide more details and I'll re-open.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug macos
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#1609
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?