github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 00:51:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #9766] Ollama 0.6.0 cannot use CUDA #6384

New Issue
Closed
opened 2026-04-12 17:53:33 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-04-12 17:53:33 -05:00
Owner
Copy Link

Originally created by @wwshs on GitHub (Mar 14, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/9766

What is the issue?

Ollama 0.6.0 for Windows.
Windows 11 Workstation.
Ollama 0.6.0 cannot use CUDA.
It can not use GPU.
Everthing use CPU.
I'v gone back to Ollama 0.5.13.

Relevant log output

OS

Windows

GPU

Nvidia

CPU

Intel

Ollama version

0.6.0

Originally created by @wwshs on GitHub (Mar 14, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/9766 ### What is the issue? Ollama 0.6.0 for Windows. Windows 11 Workstation. Ollama 0.6.0 cannot use CUDA. It can not use GPU. Everthing use CPU. I'v gone back to Ollama 0.5.13. ### Relevant log output ```shell ``` ### OS Windows ### GPU Nvidia ### CPU Intel ### Ollama version 0.6.0
GiteaMirror added the bug label 2026-04-12 17:53:33 -05:00
GiteaMirror commented 2026-04-12 17:53:36 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Mar 14, 2025):

@wwshs sorry to hear. Do you have the logs by chance?

<!-- gh-comment-id:2724906539 --> @jmorganca commented on GitHub (Mar 14, 2025): @wwshs sorry to hear. Do you have the [logs](https://github.com/user-attachments/files/18643815/server.log) by chance?
GiteaMirror commented 2026-04-12 17:53:36 -05:00
Author
Owner
Copy Link

@mehditahmasebi commented on GitHub (Mar 14, 2025):

me too

<!-- gh-comment-id:2724926556 --> @mehditahmasebi commented on GitHub (Mar 14, 2025): me too
GiteaMirror commented 2026-04-12 17:53:37 -05:00
Author
Owner
Copy Link

@Uncle-Enzo commented on GitHub (Mar 14, 2025):

I had the same issue in wsl with Gemma3:4b. But in fixing it through forcing Cuda usage I now see a lot of this:

ggml_cuda_host_malloc: failed to allocate 0.00 MiB of pinned memory: out of memory
ggml_cuda_host_malloc: failed to allocate 0.00 MiB of pinned memory: out of memory
ggml_cuda_host_malloc: failed to allocate 0.01 MiB of pinned memory: out of memory
ggml_cuda_host_malloc: failed to allocate 0.01 MiB of pinned memory: out of memory
ggml_cuda_host_malloc: failed to allocate 0.00 MiB of pinned memory: out of memory
ggml_cuda_host_malloc: failed to allocate 0.00 MiB of pinned memory: out of memory

Very odd.

<!-- gh-comment-id:2725123280 --> @Uncle-Enzo commented on GitHub (Mar 14, 2025): I had the same issue in wsl with Gemma3:4b. But in fixing it through forcing Cuda usage I now see a lot of this: ggml_cuda_host_malloc: failed to allocate 0.00 MiB of pinned memory: out of memory ggml_cuda_host_malloc: failed to allocate 0.00 MiB of pinned memory: out of memory ggml_cuda_host_malloc: failed to allocate 0.01 MiB of pinned memory: out of memory ggml_cuda_host_malloc: failed to allocate 0.01 MiB of pinned memory: out of memory ggml_cuda_host_malloc: failed to allocate 0.00 MiB of pinned memory: out of memory ggml_cuda_host_malloc: failed to allocate 0.00 MiB of pinned memory: out of memory Very odd.
GiteaMirror commented 2026-04-12 17:53:38 -05:00
Author
Owner
Copy Link

@Mohamed0Hegazi commented on GitHub (Mar 17, 2025):

https://code.superxiang.com/

<!-- gh-comment-id:2728852684 --> @Mohamed0Hegazi commented on GitHub (Mar 17, 2025): https://code.superxiang.com/
GiteaMirror commented 2026-04-12 17:53:40 -05:00
Author
Owner
Copy Link

@seanfarr788 commented on GitHub (Mar 17, 2025):

I had this issue where my install was a locally built version of ollama. I think it may have to do with a mixture of installation methods but uncertain. By running

sudo kill $(pidof ollama)

it seemed to fix it.

<!-- gh-comment-id:2730212144 --> @seanfarr788 commented on GitHub (Mar 17, 2025): I had this issue where my install was a locally built version of ollama. I think it may have to do with a mixture of installation methods but uncertain. By running `sudo kill $(pidof ollama)` it seemed to fix it.
GiteaMirror commented 2026-04-12 17:53:41 -05:00
Author
Owner
Copy Link

@molkemon commented on GitHub (Apr 3, 2025):

I am gettting the same thing. I am using a powershell script to feed several hundred text files to amodel to replace dynamic localisation with generic wording for a game mod (ie [Root.Monarch.GetName] should be replaced with "our ruler" or something like that. It works fine, but after a few hundred gens it starts spamming the couldn't allocate 0mb VRAM error.

I have 24gb and it happens with both gemma3:27b and gemma3:12b, even though with both i still have some VRAM left (a lot with the 12b model obviously).

I really don't see a reason why this would happen, I don't think I have enabled history or something (unless I have to specifically turn the api to turn it off? butt i can't find anything in documenation). Generation continues, but way slower. So I have to stop my script and then start it again. It will continue where it left off, but this is still very annoying.

<!-- gh-comment-id:2776460834 --> @molkemon commented on GitHub (Apr 3, 2025): I am gettting the same thing. I am using a powershell script to feed several hundred text files to amodel to replace dynamic localisation with generic wording for a game mod (ie [Root.Monarch.GetName] should be replaced with "our ruler" or something like that. It works fine, but after a few hundred gens it starts spamming the couldn't allocate 0mb VRAM error. I have 24gb and it happens with both gemma3:27b and gemma3:12b, even though with both i still have some VRAM left (a lot with the 12b model obviously). I really don't see a reason why this would happen, I don't think I have enabled history or something (unless I have to specifically turn the api to turn it off? butt i can't find anything in documenation). Generation continues, but way slower. So I have to stop my script and then start it again. It will continue where it left off, but this is still very annoying.
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
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#6384
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?