github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 05:05:37 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #9408] how to use slurm to run the Ollama service. #6133

New Issue
Open
opened 2026-04-12 17:28:43 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 17:28:43 -05:00
Owner
Copy Link

Originally created by @jjxyhb on GitHub (Feb 28, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/9408

I want to configure ollama on a specific node and gpu in a lenovo lico cluster Thanks

Originally created by @jjxyhb on GitHub (Feb 28, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/9408 I want to configure ollama on a specific node and gpu in a lenovo lico cluster Thanks
GiteaMirror added the question label 2026-04-12 17:28:43 -05:00
GiteaMirror commented 2026-04-12 17:28:44 -05:00
Author
Owner
Copy Link

@flywiththetide commented on GitHub (Mar 4, 2025):

You can run Ollama on a specific node and GPU in a Slurm-managed cluster by using a Slurm job script. Here’s how to do it:

1. Create a Slurm Job Script (run_ollama.sh)

Create a Slurm script to request a specific node and GPU:

#!/bin/bash
#SBATCH --job-name=ollama-job
#SBATCH --partition=gpu
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --gres=gpu:1
#SBATCH --constraint=YOUR_GPU_MODEL
#SBATCH --output=ollama_output.log
#SBATCH --error=ollama_error.log

# Load necessary modules
module load cuda/12.1

# Activate the environment if needed
source ~/.bashrc
conda activate ollama_env  # or source venv/bin/activate

# Run Ollama server
ollama serve

2. Submit the Job

Submit the script using:

sbatch run_ollama.sh

This will:

  • Request one GPU on an available node.
  • Log outputs to ollama_output.log.
  • Ensure Ollama runs with GPU acceleration.

3. Verify GPU Allocation

Once the job is running, check GPU usage with:

squeue --user=$USER
srun --jobid=<JOB_ID> nvidia-smi

Let me know if you need help adjusting this for specific GPUs or multi-node execution!

<!-- gh-comment-id:2696234241 --> @flywiththetide commented on GitHub (Mar 4, 2025): You can run Ollama on a **specific node and GPU** in a Slurm-managed cluster by using a Slurm job script. Here’s how to do it: ### **1. Create a Slurm Job Script (`run_ollama.sh`)** Create a Slurm script to request a **specific node and GPU**: ```bash #!/bin/bash #SBATCH --job-name=ollama-job #SBATCH --partition=gpu #SBATCH --nodes=1 #SBATCH --ntasks-per-node=1 #SBATCH --gres=gpu:1 #SBATCH --constraint=YOUR_GPU_MODEL #SBATCH --output=ollama_output.log #SBATCH --error=ollama_error.log # Load necessary modules module load cuda/12.1 # Activate the environment if needed source ~/.bashrc conda activate ollama_env # or source venv/bin/activate # Run Ollama server ollama serve ``` ### **2. Submit the Job** Submit the script using: ```bash sbatch run_ollama.sh ``` This will: - **Request one GPU** on an available node. - **Log outputs** to `ollama_output.log`. - **Ensure Ollama runs** with GPU acceleration. ### **3. Verify GPU Allocation** Once the job is running, check GPU usage with: ```bash squeue --user=$USER srun --jobid=<JOB_ID> nvidia-smi ``` Let me know if you need help adjusting this for **specific GPUs or multi-node execution**!
GiteaMirror commented 2026-04-12 17:28:44 -05:00
Author
Owner
Copy Link

@giuliofrey commented on GitHub (Jun 12, 2025):

I created a template that might be useful for others that come to this issue. You can find it here

<!-- gh-comment-id:2966859796 --> @giuliofrey commented on GitHub (Jun 12, 2025): I created a template that might be useful for others that come to this issue. You can find it [here](https://github.com/giuliofrey/slurm_ollama)
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
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#6133
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?