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 #7061] GPU Tesla at 100% and ollama don't work, is hung #66540

New Issue
Closed
opened 2026-05-04 07:19:24 -05:00 by GiteaMirror · 5 comments
GiteaMirror commented 2026-05-04 07:19:24 -05:00
Owner
Copy Link

Originally created by @Domi31tls on GitHub (Oct 1, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7061

Originally assigned to: @dhiltgen on GitHub.

What is the issue?

image

image

image

I have just set up my Debian 12 Linux server. I installed Ollama directly using the following command:

curl -fsSL https://ollama.com/install.sh | sh

I did not make any modifications to the service file. I ran the command:

ollama run phi3.5

I typed "hello". The response was good and correct. However, for my second question, it started writing the beginning of a sentence and then displayed "###" (see screenshot).

I ran the nvidia-smi command (see screenshot), and the processor stays at 100%, and Ollama stops responding.

I should mention that my server does not have a graphical interface, and the Nvidia and CUDA drivers are up to date (update, upgrade). I am providing the top command output for CPU and RAM information.

Thank you for your feedback.

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

0.3.12

Originally created by @Domi31tls on GitHub (Oct 1, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7061 Originally assigned to: @dhiltgen on GitHub. ### What is the issue? ![image](https://github.com/user-attachments/assets/b187c01f-bb03-4116-b7f6-1d07553af8eb) ![image](https://github.com/user-attachments/assets/55b19214-76e7-4f14-9edd-e2b739ee965f) ![image](https://github.com/user-attachments/assets/02a31135-2884-4060-8e7a-6b271a91ec7e) I have just set up my Debian 12 Linux server. I installed Ollama directly using the following command: curl -fsSL https://ollama.com/install.sh | sh I did not make any modifications to the service file. I ran the command: ollama run phi3.5 I typed "hello". The response was good and correct. However, for my second question, it started writing the beginning of a sentence and then displayed "###" (see screenshot). I ran the nvidia-smi command (see screenshot), and the processor stays at 100%, and Ollama stops responding. I should mention that my server does not have a graphical interface, and the Nvidia and CUDA drivers are up to date (update, upgrade). I am providing the top command output for CPU and RAM information. Thank you for your feedback. ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version 0.3.12
GiteaMirror added the bugnvidianeeds more info labels 2026-05-04 07:19:25 -05:00
GiteaMirror commented 2026-05-04 07:19:32 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 1, 2024):

Server logs will aid in debugging.

<!-- gh-comment-id:2386332189 --> @rick-github commented on GitHub (Oct 1, 2024): [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) will aid in debugging.
GiteaMirror commented 2026-05-04 07:19:33 -05:00
Author
Owner
Copy Link

@Domi31tls commented on GitHub (Oct 1, 2024):

Thank for your help
debug.txt

<!-- gh-comment-id:2386674520 --> @Domi31tls commented on GitHub (Oct 1, 2024): Thank for your help [debug.txt](https://github.com/user-attachments/files/17214266/debug.txt)
GiteaMirror commented 2026-05-04 07:19:34 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Oct 2, 2024):

Unfortunately there's nothing untoward in the log. The model loads successfully, your bonjour is processed in 181ms and your question apparently in 129s. There are some other commands I take to be ollama ps and ollama list. The only curious entries are some /api/generate commands that are forced to complete when you took the service down. These might have been attempts to unload/reload the model but it's not clear.

You can improve the logging by adding OLLAMA_DEBUG=1 to the server environment. And since the card is a few years old, you could try using an older version of the CUDA library by setting OLLAMA_LLM_LIBRARY=cuda_v11.

<!-- gh-comment-id:2387452139 --> @rick-github commented on GitHub (Oct 2, 2024): Unfortunately there's nothing untoward in the log. The model loads successfully, your `bonjour` is processed in 181ms and your question apparently in 129s. There are some other commands I take to be `ollama ps` and `ollama list`. The only curious entries are some `/api/generate` commands that are forced to complete when you took the service down. These might have been attempts to unload/reload the model but it's not clear. You can improve the logging by adding `OLLAMA_DEBUG=1` to the [server environment](https://github.com/ollama/ollama/blob/main/docs/faq.md#setting-environment-variables-on-linux). And since the card is a few years old, you could try using an older version of the CUDA library by setting `OLLAMA_LLM_LIBRARY=cuda_v11`.
GiteaMirror commented 2026-05-04 07:19:34 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Nov 5, 2024):

@Domi31tls can you try out 0.4.0 and see if that improves the situation? Do you see this failure with other models, or just phi3.5?

<!-- gh-comment-id:2458315501 --> @dhiltgen commented on GitHub (Nov 5, 2024): @Domi31tls can you try out 0.4.0 and see if that improves the situation? Do you see this failure with other models, or just phi3.5?
GiteaMirror commented 2026-05-04 07:19:35 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Jan 4, 2026):

Will close for now since we need some more info, but feel free to let us know if you're still hitting this.

<!-- gh-comment-id:3707543376 --> @jmorganca commented on GitHub (Jan 4, 2026): Will close for now since we need some more info, but feel free to let us know if you're still hitting this.
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 nvidia
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#66540
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?