github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-05 23:53:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #4977] qwen2-72b start to output gibberish at some point if i set num_ctx to 8192 #3146

New Issue
Open
opened 2026-04-12 13:37:29 -05:00 by GiteaMirror · 4 comments
GiteaMirror commented 2026-04-12 13:37:29 -05:00
Owner
Copy Link

Originally created by @Mikhael-Danilov on GitHub (Jun 11, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4977

What is the issue?

qwen2-72b start to output gibberish like this:
.5"5.F9(CB;6@FC9!DC:$B$D60G5",3B+2;1-*,@%=876E0;5*:.98G4!980+D
at some point if i set num_ctx to 8192.
Normal output from llm was expected.

Issue persist when using ollama run, or when using api (Silly Tavern)

qwen2-72b works fine with num_ctx 2048

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

0.1.42

Originally created by @Mikhael-Danilov on GitHub (Jun 11, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4977 ### What is the issue? qwen2-72b start to output gibberish like this: `.5"5.F9(CB;6@FC9!DC:$B$D60G5",3B+2;1-*,@%=876E0;5*:.98G4!980+D` at some point if i set num_ctx to 8192. Normal output from llm was expected. Issue persist when using `ollama run`, or when using api (Silly Tavern) qwen2-72b works fine with num_ctx 2048 ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version 0.1.42
GiteaMirror added the bug label 2026-04-12 13:37:29 -05:00
GiteaMirror commented 2026-04-12 13:37:30 -05:00
Author
Owner
Copy Link

@kac487 commented on GitHub (Jul 8, 2024):

I am seeing this as well. I am unable to get this model to output anything coherent with longer context lengths

<!-- gh-comment-id:2213052993 --> @kac487 commented on GitHub (Jul 8, 2024): I am seeing this as well. I am unable to get this model to output anything coherent with longer context lengths
GiteaMirror commented 2026-04-12 13:37:31 -05:00
Author
Owner
Copy Link

@Mikhael-Danilov commented on GitHub (Jul 9, 2024):

Issue still persists in version 0.2.1

<!-- gh-comment-id:2218737258 --> @Mikhael-Danilov commented on GitHub (Jul 9, 2024): Issue still persists in version 0.2.1
GiteaMirror commented 2026-04-12 13:37:32 -05:00
Author
Owner
Copy Link

@haining78zhang commented on GitHub (Aug 27, 2024):

any updates, issue still there version 0.3.6

<!-- gh-comment-id:2311860158 --> @haining78zhang commented on GitHub (Aug 27, 2024): any updates, issue still there version 0.3.6
GiteaMirror commented 2026-04-12 13:37:34 -05:00
Author
Owner
Copy Link

@haining78zhang commented on GitHub (Aug 27, 2024):

but if I use Qwen2 7b, it runs well, if I use qwen2-72b-instruct-q4_0.gguf a self imported model, I got this issue

<!-- gh-comment-id:2311863373 --> @haining78zhang commented on GitHub (Aug 27, 2024): but if I use Qwen2 7b, it runs well, if I use qwen2-72b-instruct-q4_0.gguf a self imported model, I got this issue
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#3146
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?