github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #7013] Option to Override a Model's Memory Requirements #66499

New Issue
Closed
opened 2026-05-04 06:51:57 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-05-04 06:51:57 -05:00
Owner
Copy Link

Originally created by @dabockster on GitHub (Sep 28, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7013

Originally assigned to: @dhiltgen on GitHub.

I was trying to load the 70b Llama3 model and Ollama says I need 33.6 GB of 30.5 GB ram. I believe this is a safety thing Meta put into the model, so I want to have the ability to override this and attempt to run it on lower amounts of memory. I know this will likely dip into swap/page file space, possibly even causing kernel panics and BSODs. But I still want to make the attempt.

Computer, disable holodeck safety protocols. Authorization Picard-4-7-Alpha-Tango.

Originally created by @dabockster on GitHub (Sep 28, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7013 Originally assigned to: @dhiltgen on GitHub. I was trying to load the 70b Llama3 model and Ollama says I need 33.6 GB of 30.5 GB ram. I believe this is a safety thing Meta put into the model, so I want to have the ability to override this and attempt to run it on lower amounts of memory. I know this will likely dip into swap/page file space, possibly even causing kernel panics and BSODs. But I still want to make the attempt. _Computer, disable holodeck safety protocols. Authorization Picard-4-7-Alpha-Tango._
GiteaMirror added the questionwindows labels 2026-05-04 06:51:58 -05:00
GiteaMirror commented 2026-05-04 06:52:18 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Sep 28, 2024):

We take swap space into consideration. If you increase the size of your swap file it should try to load it, but expect extremely slow performance if you're paging in/out of swap to fit the model.

If that doesn't work, please share your server log so we can see more details on why it wouldn't load the model and I'll reopen the issue.

<!-- gh-comment-id:2381015703 --> @dhiltgen commented on GitHub (Sep 28, 2024): We take swap space into consideration. If you increase the size of your swap file it should try to load it, but expect extremely slow performance if you're paging in/out of swap to fit the model. If that doesn't work, please share your server log so we can see more details on why it wouldn't load the model and I'll reopen the issue.
GiteaMirror commented 2026-05-04 06:52:20 -05:00
Author
Owner
Copy Link

@ndrsvh commented on GitHub (Jan 9, 2025):

We take swap space into consideration. If you increase the size of your swap file it should try to load it, but expect extremely slow performance if you're paging in/out of swap to fit the model.

If that doesn't work, please share your server log so we can see more details on why it wouldn't load the model and I'll reopen the issue.

Hello! Could you clarify, I have multiple swap files on different drives, which file will be used by ollama? Do I understand correctly that the one that will be determined by the operating system itself?

<!-- gh-comment-id:2580325393 --> @ndrsvh commented on GitHub (Jan 9, 2025): > We take swap space into consideration. If you increase the size of your swap file it should try to load it, but expect extremely slow performance if you're paging in/out of swap to fit the model. > > If that doesn't work, please share your server log so we can see more details on why it wouldn't load the model and I'll reopen the issue. Hello! Could you clarify, I have multiple swap files on different drives, which file will be used by ollama? Do I understand correctly that the one that will be determined by the operating system itself?
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 question windows
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#66499
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?