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 #7576] num_ctx causes 100% CPU with no GPU usage #51340

New Issue
Closed
opened 2026-04-28 19:35:35 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-04-28 19:35:35 -05:00
Owner
Copy Link

Originally created by @aaronbolton on GitHub (Nov 8, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7576

What is the issue?

ive tried to create a new model recently with only the parameter num_ctx, when I run the models it shows 100% CPU with no GPU usage, even if the model was too big I would assume it would report GPU/CPU 100%/???%

OS

Docker

GPU

Nvidia

CPU

Intel

Ollama version

0.4.0

Originally created by @aaronbolton on GitHub (Nov 8, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7576 ### What is the issue? ive tried to create a new model recently with only the parameter num_ctx, when I run the models it shows 100% CPU with no GPU usage, even if the model was too big I would assume it would report GPU/CPU 100%/???% ### OS Docker ### GPU Nvidia ### CPU Intel ### Ollama version 0.4.0
GiteaMirror added the bugneeds more info labels 2026-04-28 19:35:36 -05:00
GiteaMirror commented 2026-04-28 19:35:41 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Nov 8, 2024):

#7509

Server logs will help in debugging.

<!-- gh-comment-id:2465216823 --> @rick-github commented on GitHub (Nov 8, 2024): #7509 [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) will help in debugging.
GiteaMirror commented 2026-04-28 19:35:42 -05:00
Author
Owner
Copy Link

@aaronbolton commented on GitHub (Nov 10, 2024):

Here is a fairly clean log, the model is loaded at around line 47

ollama_logs.txt

<!-- gh-comment-id:2466707819 --> @aaronbolton commented on GitHub (Nov 10, 2024): Here is a fairly clean log, the model is loaded at around line 47 [ollama_logs.txt](https://github.com/user-attachments/files/17691317/ollama_logs.txt)
GiteaMirror commented 2026-04-28 19:35:42 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Nov 10, 2024):

You have OLLAMA_NUM_PARALLEL=16 and are setting num_ctx to 32K, so the runner needs 28G just for context. Combined with other memory requirements, it just won't fit in the 2x15G you have on your GPUs, so the entire model is loaded in RAM. Try reducing OLLAMA_NUM_PARALLEL or num_ctx.

<!-- gh-comment-id:2466721302 --> @rick-github commented on GitHub (Nov 10, 2024): You have `OLLAMA_NUM_PARALLEL=16` and are setting `num_ctx` to 32K, so the runner needs 28G just for context. Combined with other memory requirements, it just won't fit in the 2x15G you have on your GPUs, so the entire model is loaded in RAM. Try reducing `OLLAMA_NUM_PARALLEL` or `num_ctx`.
GiteaMirror commented 2026-04-28 19:35:43 -05:00
Author
Owner
Copy Link

@aaronbolton commented on GitHub (Nov 12, 2024):

perfect thank you for the pointer and informative answer

<!-- gh-comment-id:2471300965 --> @aaronbolton commented on GitHub (Nov 12, 2024): perfect thank you for the pointer and informative answer
GiteaMirror commented 2026-04-28 19:35:44 -05:00
Author
Owner
Copy Link

@liaoweiguo commented on GitHub (Dec 29, 2024):

how do we get detailed information about this memory information, so important

<!-- gh-comment-id:2564624949 --> @liaoweiguo commented on GitHub (Dec 29, 2024): how do we get detailed information about this memory information, so important
GiteaMirror commented 2026-04-28 19:35:45 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Dec 29, 2024):

Server logs. Look for memory.required.

<!-- gh-comment-id:2564626484 --> @rick-github commented on GitHub (Dec 29, 2024): [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues). Look for `memory.required`.
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 needs more info
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#51340
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?