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 #14720] Running deepseek-r1:32b doesn't use my 3090s for inferencing anymore #56030

New Issue
Closed
opened 2026-04-29 10:09:50 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-29 10:09:50 -05:00
Owner
Copy Link

Originally created by @alpha754293 on GitHub (Mar 8, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/14720

What is the issue?

I am running the deepseek-r1:32b model and it loaded the LLM into the VRAM split between my two 3090s, but now, it doesn't use neither of them for the actual inferencing/responding part.

I am not sure what changed nor when it changed, but I just noticed this behaviour today.

Relevant log output

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

0.17.7

Originally created by @alpha754293 on GitHub (Mar 8, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/14720 ### What is the issue? I am running the `deepseek-r1:32b` model and it loaded the LLM into the VRAM split between my two 3090s, but now, it doesn't use neither of them for the actual inferencing/responding part. I am not sure what changed nor when it changed, but I just noticed this behaviour today. ### Relevant log output ```shell ``` ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version 0.17.7
GiteaMirror added the bug label 2026-04-29 10:09:50 -05:00
GiteaMirror commented 2026-04-29 10:09:52 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Mar 8, 2026):

Perhaps #14116. Server logs will aid in debugging.

<!-- gh-comment-id:4019680658 --> @rick-github commented on GitHub (Mar 8, 2026): Perhaps #14116. [Server logs](https://docs.ollama.com/troubleshooting) will aid in debugging.
GiteaMirror commented 2026-04-29 10:09:52 -05:00
Author
Owner
Copy Link

@alpha754293 commented on GitHub (Mar 8, 2026):

Perhaps #14116. Server logs will aid in debugging.

My apologies. Here you go.

log.txt

And here is the docker compose YAML file that I use (inside dockge) to start it, but it seems to be mostly different to how I am starting ollama.

compose.yaml

<!-- gh-comment-id:4019715045 --> @alpha754293 commented on GitHub (Mar 8, 2026): > Perhaps [#14116](https://github.com/ollama/ollama/issues/14116). [Server logs](https://docs.ollama.com/troubleshooting) will aid in debugging. My apologies. Here you go. [log.txt](https://github.com/user-attachments/files/25826699/log.txt) And here is the docker compose YAML file that I use (inside dockge) to start it, but it seems to be mostly different to how I am starting ollama. [compose.yaml](https://github.com/user-attachments/files/25826716/compose.yaml)
GiteaMirror commented 2026-04-29 10:09:52 -05:00
Author
Owner
Copy Link

@alpha754293 commented on GitHub (Mar 8, 2026):

I think that I might have figured it out. It would appear that as of v0.15.5, there was a change to the context window defaults that for systems with a total of 48 GB of VRAM (I presume), that it set the default context window size to 32768.

Thanks.

<!-- gh-comment-id:4019751585 --> @alpha754293 commented on GitHub (Mar 8, 2026): I think that I might have figured it out. It would appear that as of v0.15.5, there was a change to the context window defaults that for systems with a total of 48 GB of VRAM (I presume), that it set the default context window size to 32768. Thanks.
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#56030
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?