github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #11697] Memory access fault with gpt-oss:20b #33501

New Issue
Closed
opened 2026-04-22 16:15:48 -05:00 by GiteaMirror · 5 comments
GiteaMirror commented 2026-04-22 16:15:48 -05:00
Owner
Copy Link

Originally created by @Expro on GitHub (Aug 5, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/11697

What is the issue?

Prompting gpt-oss:20b on 0.11.1 results in error:

Memory access fault by GPU node-1 (Agent handle: 0x7f3b0c66f890) on address 0x7f360519e000. Reason: Page not present or supervisor privilege.

Relevant log output

time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:367 msg="offloading 24 repeating layers to GPU"
time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:373 msg="offloading output layer to GPU"
time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:378 msg="offloaded 25/25 layers to GPU"
time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:381 msg="model weights" buffer=ROCm0 size="11.7 GiB"
time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:381 msg="model weights" buffer=CPU size="1.1 GiB"
time=2025-08-05T20:35:49.188Z level=INFO source=ggml.go:672 msg="compute graph" backend=ROCm0 buffer_type=ROCm0 size="2.1 GiB"
time=2025-08-05T20:35:49.188Z level=INFO source=ggml.go:672 msg="compute graph" backend=CPU buffer_type=CPU size="5.6 MiB"
time=2025-08-05T20:36:11.960Z level=INFO source=server.go:637 msg="llama runner started in 25.11 seconds"
Memory access fault by GPU node-1 (Agent handle: 0x7fedb466f890) on address 0x7fec0d423000. Reason: Page not present or supervisor privilege.
time=2025-08-05T20:36:34.898Z level=ERROR source=server.go:807 msg="post predict" error="Post \"http://127.0.0.1:33281/completion\": EOF"
[GIN] 2025/08/05 - 20:36:34 | 200 | 48.888797725s |      10.244.2.0 | POST     "/api/chat"
time=2025-08-05T20:36:35.320Z level=WARN source=server.go:517 msg="llama runner process no longer running" sys=134 string="signal: aborted (core dumped)"

OS

Docker

GPU

AMD

CPU

AMD

Ollama version

0.11.1

Originally created by @Expro on GitHub (Aug 5, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/11697 ### What is the issue? Prompting gpt-oss:20b on 0.11.1 results in error: Memory access fault by GPU node-1 (Agent handle: 0x7f3b0c66f890) on address 0x7f360519e000. Reason: Page not present or supervisor privilege. ### Relevant log output ```shell time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:367 msg="offloading 24 repeating layers to GPU" time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:373 msg="offloading output layer to GPU" time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:378 msg="offloaded 25/25 layers to GPU" time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:381 msg="model weights" buffer=ROCm0 size="11.7 GiB" time=2025-08-05T20:35:49.174Z level=INFO source=ggml.go:381 msg="model weights" buffer=CPU size="1.1 GiB" time=2025-08-05T20:35:49.188Z level=INFO source=ggml.go:672 msg="compute graph" backend=ROCm0 buffer_type=ROCm0 size="2.1 GiB" time=2025-08-05T20:35:49.188Z level=INFO source=ggml.go:672 msg="compute graph" backend=CPU buffer_type=CPU size="5.6 MiB" time=2025-08-05T20:36:11.960Z level=INFO source=server.go:637 msg="llama runner started in 25.11 seconds" Memory access fault by GPU node-1 (Agent handle: 0x7fedb466f890) on address 0x7fec0d423000. Reason: Page not present or supervisor privilege. time=2025-08-05T20:36:34.898Z level=ERROR source=server.go:807 msg="post predict" error="Post \"http://127.0.0.1:33281/completion\": EOF" [GIN] 2025/08/05 - 20:36:34 | 200 | 48.888797725s | 10.244.2.0 | POST "/api/chat" time=2025-08-05T20:36:35.320Z level=WARN source=server.go:517 msg="llama runner process no longer running" sys=134 string="signal: aborted (core dumped)" ``` ### OS Docker ### GPU AMD ### CPU AMD ### Ollama version 0.11.1
GiteaMirror added the bug label 2026-04-22 16:15:48 -05:00
GiteaMirror commented 2026-04-22 16:15:51 -05:00
Author
Owner
Copy Link

@dtori commented on GitHub (Aug 5, 2025):

Same
$ ollama run gpt-oss:20b

Hello
Error: model runner has unexpectedly stopped, this may be due to resource limitations or an internal error, check ollama server logs for details

(ubuntu linux, rtx3090)

<!-- gh-comment-id:3156713476 --> @dtori commented on GitHub (Aug 5, 2025): Same $ ollama run gpt-oss:20b >>> Hello Error: model runner has unexpectedly stopped, this may be due to resource limitations or an internal error, check ollama server logs for details (ubuntu linux, rtx3090)
GiteaMirror commented 2026-04-22 16:15:52 -05:00
Author
Owner
Copy Link

@ayylmaonade commented on GitHub (Aug 5, 2025):

+1
Having the same issue here.
I found a temporary workaround: You have to disable flash attention. Just serve ollama like this; `OLLAMA_FLASH_ATTENTION=0 ollama serve** until they solve it. 7900 XTX, Arch Linux, Kernel 6.15.9.

<!-- gh-comment-id:3156867376 --> @ayylmaonade commented on GitHub (Aug 5, 2025): +1 Having the same issue here. **I found a temporary workaround**: You have to disable flash attention. Just serve ollama like this; `OLLAMA_FLASH_ATTENTION=0 ollama serve** until they solve it. 7900 XTX, Arch Linux, Kernel 6.15.9.
GiteaMirror commented 2026-04-22 16:15:53 -05:00
Author
Owner
Copy Link

@AlKhrulev commented on GitHub (Aug 5, 2025):

+1 Same problem as @dtori on M4 Pro 48 Gb with Ollama 0.11.0 running as a brew service

<!-- gh-comment-id:3156877096 --> @AlKhrulev commented on GitHub (Aug 5, 2025): +1 Same problem as @dtori on M4 Pro 48 Gb with Ollama 0.11.0 running as a brew service
GiteaMirror commented 2026-04-22 16:15:54 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (Aug 6, 2025):

This should be fixed in 0.11.2.

<!-- gh-comment-id:3157076599 --> @pdevine commented on GitHub (Aug 6, 2025): This should be fixed in `0.11.2`.
GiteaMirror commented 2026-04-22 16:15:55 -05:00
Author
Owner
Copy Link

@AlKhrulev commented on GitHub (Aug 6, 2025):

This should be fixed in 0.11.2.

yup, can confirm that it is working now. Thanks!

<!-- gh-comment-id:3157088417 --> @AlKhrulev commented on GitHub (Aug 6, 2025): > This should be fixed in `0.11.2`. yup, can confirm that it is working now. Thanks!
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
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#33501
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?