[GH-ISSUE #7363] Default CPU or GPU for Models #4679

Closed
opened 2026-04-12 15:36:16 -05:00 by GiteaMirror · 1 comment
Owner

Originally created by @makewrks on GitHub (Oct 25, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7363

Hi there,

It would be nice to see a feature that allows users to set default CPU or GPU usage for specific models in Ollama. For instance, a 1.5 code completion model could default to CPU, while larger chatbot models use the GPU.

Additionally, having the option to set specific sleep times for each model would help with resource management.

Thank you

Originally created by @makewrks on GitHub (Oct 25, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7363 Hi there, It would be nice to see a feature that allows users to set default CPU or GPU usage for specific models in Ollama. For instance, a 1.5 code completion model could default to CPU, while larger chatbot models use the GPU. Additionally, having the option to set specific sleep times for each model would help with resource management. Thank you
GiteaMirror added the feature request label 2026-04-12 15:36:16 -05:00
Author
Owner

@rick-github commented on GitHub (Oct 25, 2024):

You can make a model load on the CPU by setting num_gpu to zero: https://github.com/ollama/ollama/issues/6950#issuecomment-2373663650. Work on managing where models are allocated is being tracked in #3902 but there's no progress so far.

<!-- gh-comment-id:2438687449 --> @rick-github commented on GitHub (Oct 25, 2024): You can make a model load on the CPU by setting `num_gpu` to zero: https://github.com/ollama/ollama/issues/6950#issuecomment-2373663650. Work on managing where models are allocated is being tracked in #3902 but there's no progress so far.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#4679