github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 08:02:14 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #12547] When I use the fine tuned model to use ollama inference, there may be incorrect answers. #8324

New Issue
Open
opened 2026-04-12 20:53:42 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 20:53:42 -05:00
Owner
Copy Link

Originally created by @zhangnn520 on GitHub (Oct 9, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12547

What is the issue?

I have a question. When I use the fine tuned model to use ollama inference, there may be incorrect answers. Generally, both vllm and sglang can be used for normal inference. However, when I use ollama without quantification, the same problem still occurs. How should I solve this problem

Image

Relevant log output

OS

No response

GPU

No response

CPU

No response

Ollama version

No response

Originally created by @zhangnn520 on GitHub (Oct 9, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12547 ### What is the issue? I have a question. When I use the fine tuned model to use ollama inference, there may be incorrect answers. Generally, both vllm and sglang can be used for normal inference. However, when I use ollama without quantification, the same problem still occurs. How should I solve this problem <img width="2423" height="895" alt="Image" src="https://github.com/user-attachments/assets/a1c16fc0-767f-43da-9a33-a52201bbe57a" /> ### Relevant log output ```shell ``` ### OS _No response_ ### GPU _No response_ ### CPU _No response_ ### Ollama version _No response_
GiteaMirror added the bug label 2026-04-12 20:53:42 -05:00
GiteaMirror commented 2026-04-12 20:53:44 -05:00
Author
Owner
Copy Link

@zhangnn520 commented on GitHub (Oct 9, 2025):

As shown in the figure, the first one using the official qwen3-0.6/1.7/4b can output normally, but when I use the qwen2.5-vl-code model, ggg occurs. The qwen2.5-vl-code model is my fine tuned model, but it seems to not work properly.

<!-- gh-comment-id:3384543991 --> @zhangnn520 commented on GitHub (Oct 9, 2025): As shown in the figure, the first one using the official qwen3-0.6/1.7/4b can output normally, but when I use the qwen2.5-vl-code model, ggg occurs. The qwen2.5-vl-code model is my fine tuned model, but it seems to not work properly.
GiteaMirror commented 2026-04-12 20:53:45 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (Oct 10, 2025):

You can compare the parameters/tensors with ollama show -v <model>. Maybe output your model against the base model to see if there are any differences.

<!-- gh-comment-id:3392242326 --> @pdevine commented on GitHub (Oct 10, 2025): You can compare the parameters/tensors with `ollama show -v <model>`. Maybe output your model against the base model to see if there are any differences.
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#8324
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?