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 #6250] 运行glm4-9b模型,对话时间久了会偶发性的回复GGGGGGG #50421

New Issue
Closed
opened 2026-04-28 15:45:25 -05:00 by GiteaMirror · 7 comments
GiteaMirror commented 2026-04-28 15:45:25 -05:00
Owner
Copy Link

Originally created by @MdcGIt on GitHub (Aug 8, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6250

Originally assigned to: @jmorganca on GitHub.

What is the issue?

运行glm4-9b模型,对话时间久了会偶发性的回复GGGGGGG
显卡信息如下:
image

OS

Linux

GPU

Intel

CPU

Intel

Ollama version

0.3.0

Originally created by @MdcGIt on GitHub (Aug 8, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6250 Originally assigned to: @jmorganca on GitHub. ### What is the issue? 运行glm4-9b模型,对话时间久了会偶发性的回复GGGGGGG 显卡信息如下: ![image](https://github.com/user-attachments/assets/5ae71376-3667-4d65-8603-549768fccb40) ### OS Linux ### GPU Intel ### CPU Intel ### Ollama version 0.3.0
GiteaMirror added the needs more infobug labels 2026-04-28 15:45:27 -05:00
GiteaMirror commented 2026-04-28 15:45:32 -05:00
Author
Owner
Copy Link

@wszgrcy commented on GitHub (Aug 8, 2024):

不一定运行久,有时候第一次正常,再来一次就失败了

<!-- gh-comment-id:2275043542 --> @wszgrcy commented on GitHub (Aug 8, 2024): 不一定运行久,有时候第一次正常,再来一次就失败了
GiteaMirror commented 2026-04-28 15:45:34 -05:00
Author
Owner
Copy Link

@AeneasZhu commented on GitHub (Aug 8, 2024):

This might be llama.cpp's problem that causes the bug. I suggest you post your issue in llama.cpp.

<!-- gh-comment-id:2276013002 --> @AeneasZhu commented on GitHub (Aug 8, 2024): This might be `llama.cpp`'s problem that causes the bug. I suggest you post your issue in `llama.cpp`.
GiteaMirror commented 2026-04-28 15:45:36 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Sep 2, 2024):

Having trouble reproducing this one. May I ask what prompt you used and which glm4 model? Was it ollama run glm4?

<!-- gh-comment-id:2325373894 --> @jmorganca commented on GitHub (Sep 2, 2024): Having trouble reproducing this one. May I ask what prompt you used and which glm4 model? Was it `ollama run glm4`?
GiteaMirror commented 2026-04-28 15:45:38 -05:00
Author
Owner
Copy Link

@wszgrcy commented on GitHub (Sep 2, 2024):

Having trouble reproducing this one. May I ask what prompt you used and which glm4 model? Was it ollama run glm4?

Llama.cpp seems to have fixed this issue
https://github.com/ggerganov/llama.cpp/pull/9130
But the code is not synchronized?

<!-- gh-comment-id:2325394129 --> @wszgrcy commented on GitHub (Sep 2, 2024): > Having trouble reproducing this one. May I ask what prompt you used and which glm4 model? Was it `ollama run glm4`? Llama.cpp seems to have fixed this issue https://github.com/ggerganov/llama.cpp/pull/9130 But the code is not synchronized?
GiteaMirror commented 2026-04-28 15:45:41 -05:00
Author
Owner
Copy Link

@shuye-cheung commented on GitHub (Sep 24, 2024):

I use

ollama run glm4

namely 9b-chat-q4_0

Then, the model will return "GGGGGGGGGGGGGGGGGGG" error。

<!-- gh-comment-id:2371530802 --> @shuye-cheung commented on GitHub (Sep 24, 2024): I use ``` ollama run glm4 ``` namely **9b-chat-q4_0** Then, the model will return "GGGGGGGGGGGGGGGGGGG" error。
GiteaMirror commented 2026-04-28 15:45:44 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Sep 30, 2024):

Please upgrade to the latest version of Ollama which includes upstream fixes for glm4 in llama.cpp. If you're still having troubles, share an updated server log and we'll reopen the issue.

<!-- gh-comment-id:2384320304 --> @dhiltgen commented on GitHub (Sep 30, 2024): Please upgrade to the latest version of Ollama which includes upstream fixes for glm4 in llama.cpp. If you're still having troubles, share an updated server log and we'll reopen the issue.
GiteaMirror commented 2026-04-28 15:45:45 -05:00
Author
Owner
Copy Link

@1003663050 commented on GitHub (Apr 15, 2025):

要看显卡支持的指令集，比如：NVIDIA GeForce RTX 3080不支持INT4指令集‌。RTX 3080使用的是Ampere架构，该架构支持INT8和FP16运算，但不直接支持INT4运算‌，但是Q4模型是需要int4指令集的

<!-- gh-comment-id:2803904972 --> @1003663050 commented on GitHub (Apr 15, 2025): 要看显卡支持的指令集，比如：NVIDIA GeForce RTX 3080不支持INT4指令集‌。RTX 3080使用的是Ampere架构，该架构支持INT8和FP16运算，但不直接支持INT4运算‌，但是Q4模型是需要int4指令集的
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#50421
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?