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 #10172] tensor-split problem #68731

New Issue
Closed
opened 2026-05-04 15:01:27 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-05-04 15:01:27 -05:00
Owner
Copy Link

Originally created by @taikai-zz on GitHub (Apr 8, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10172

cmd="/usr/local/bin/ollama runner --model /usr/share/ollama/.ollama/models/blobs/sha256-e765360bed4cbcc829a85dcb7e7cfff4fd0461a3210b555bec6bd0d2faf27b75 --ctx-size 2048 --batch-size 512 --n-gpu-layers 66 --verbose --threads 32 --parallel 1 --tensor-split 8,8,8,8,8,8,7,7 --port 44979"

This is my debugging information. I found that ollama automatically starts with the “--tensor-split” parameter set. Can I manually or set this parameter in the configuration file? I couldn't find any relevant documents
For example, I currently have a 4G sized model and I want it to run on two graphics cards.

Originally created by @taikai-zz on GitHub (Apr 8, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10172 cmd="/usr/local/bin/ollama runner --model /usr/share/ollama/.ollama/models/blobs/sha256-e765360bed4cbcc829a85dcb7e7cfff4fd0461a3210b555bec6bd0d2faf27b75 --ctx-size 2048 --batch-size 512 --n-gpu-layers 66 --verbose --threads 32 --parallel 1 --tensor-split 8,8,8,8,8,8,7,7 --port 44979" This is my debugging information. I found that ollama automatically starts with the “--tensor-split” parameter set. Can I manually or set this parameter in the configuration file? I couldn't find any relevant documents For example, I currently have a 4G sized model and I want it to run on two graphics cards.
GiteaMirror added the feature request label 2026-05-04 15:01:27 -05:00
GiteaMirror commented 2026-05-04 15:01:30 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Apr 8, 2025):

Currently you can't control GPU assignment on a per-model basis. If you want to use a subset of the available cards to run a model, you can start another server and bind specific cards to it by setting the CUDA_VISIBLE_DEVICES environment variable.

<!-- gh-comment-id:2784986895 --> @rick-github commented on GitHub (Apr 8, 2025): Currently you can't control GPU assignment on a per-model basis. If you want to use a subset of the available cards to run a model, you can start another server and bind specific cards to it by setting the `CUDA_VISIBLE_DEVICES` environment variable.
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 feature request
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#68731
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?