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 #12329] Modelfile for qwen3:4b is not working #8191

New Issue
Closed
opened 2026-04-12 20:38:29 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-12 20:38:29 -05:00
Owner
Copy Link

Originally created by @redaachouhad on GitHub (Sep 18, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12329

What is the issue?

i have this modelfile to create a finetunned model (in .gguf format) based on the qwen3:4 model, but the problem is that the tool calling is not work ? does anyone know what should be the correct template to deploy my model in ollama locally?

this my actual model file:
FROM qwen3:4b

SYSTEM """You are an assistant that discuss with customers in a frinedly way"""
TEMPLATE """
{{ if .System }}
<|im_start|>system
{{ .System }}

{{ if .Tools }}

Tools

You may call one or more functions to assist with the user query.
You are provided with function signatures within XML tags:

{{ range .Tools }} {"type": "function", "function": {"name":"{{ .Function.Name }}","description":"{{ .Function.Description }}","parameters":{{ .Function.Parameters }}}} {{ end }} {{ end }} <|im_end|> {{ end }}

{{ range $i, $msg := .Messages }}

{{ if eq $msg.Role "user" }}
<|im_start|>user
{{ $msg.Content }}
<|im_end|>

{{ else if eq $msg.Role "assistant" }}
<|im_start|>assistant
{{ if $msg.Content }}{{ $msg.Content }}{{ end }}
{{ if $msg.ToolCalls }}
{{ range $msg.ToolCalls }}
{{/* This signals Ollama to create a tool call, not just text */}}
<tool_call>
{"name": "{{ .Function.Name }}", "arguments": {{ .Function.Arguments }}}
</tool_call>
{{ end }}
{{ end }}
<|im_end|>

{{ else if eq $msg.Role "tool" }}
<|im_start|>user
<tool_response>
{{ $msg.Content }}
</tool_response>
<|im_end|>
{{ end }}

{{ end }}
"""

PARAMETER stop "<|im_start|>"
PARAMETER stop "<|im_end|>"

Relevant log output

OS

No response

GPU

Nvidia

CPU

No response

Ollama version

No response

Originally created by @redaachouhad on GitHub (Sep 18, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12329 ### What is the issue? i have this modelfile to create a finetunned model (in .gguf format) based on the qwen3:4 model, but the problem is that the tool calling is not work ? does anyone know what should be the correct template to deploy my model in ollama locally? this my actual model file: FROM qwen3:4b SYSTEM """You are an assistant that discuss with customers in a frinedly way""" TEMPLATE """ {{ if .System }} <|im_start|>system {{ .System }} {{ if .Tools }} # Tools You may call one or more functions to assist with the user query. You are provided with function signatures within <tools></tools> XML tags: <tools> {{ range .Tools }} {"type": "function", "function": {"name":"{{ .Function.Name }}","description":"{{ .Function.Description }}","parameters":{{ .Function.Parameters }}}} {{ end }} </tools> {{ end }} <|im_end|> {{ end }} {{ range $i, $msg := .Messages }} {{ if eq $msg.Role "user" }} <|im_start|>user {{ $msg.Content }} <|im_end|> {{ else if eq $msg.Role "assistant" }} <|im_start|>assistant {{ if $msg.Content }}<think>{{ $msg.Content }}</think>{{ end }} {{ if $msg.ToolCalls }} {{ range $msg.ToolCalls }} {{/* This signals Ollama to create a tool call, not just text */}} <tool_call> {"name": "{{ .Function.Name }}", "arguments": {{ .Function.Arguments }}} </tool_call> {{ end }} {{ end }} <|im_end|> {{ else if eq $msg.Role "tool" }} <|im_start|>user <tool_response> {{ $msg.Content }} </tool_response> <|im_end|> {{ end }} {{ end }} """ PARAMETER stop "<|im_start|>" PARAMETER stop "<|im_end|>" ### Relevant log output ```shell ``` ### OS _No response_ ### GPU Nvidia ### CPU _No response_ ### Ollama version _No response_
GiteaMirror added the bug label 2026-04-12 20:38:29 -05:00
GiteaMirror commented 2026-04-12 20:38:31 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 18, 2025):

If it's a finetune of qwen3:4b, just use the qwen3:4b modelfile.

echo FROM path/to/my/model.gguf > Modelfile
echo SYSTEM '"""You are an assistant that discusses with customers in a friendly way"""' >> Modelfile
ollama show --modelfile qwen3:4b | grep -v "^FROM" >> Modelfile
ollama create my-new-model
<!-- gh-comment-id:3306711926 --> @rick-github commented on GitHub (Sep 18, 2025): If it's a finetune of qwen3:4b, just use the qwen3:4b modelfile. ``` echo FROM path/to/my/model.gguf > Modelfile echo SYSTEM '"""You are an assistant that discusses with customers in a friendly way"""' >> Modelfile ollama show --modelfile qwen3:4b | grep -v "^FROM" >> Modelfile ollama create my-new-model ```
GiteaMirror commented 2026-04-12 20:38:32 -05:00
Author
Owner
Copy Link

@redaachouhad commented on GitHub (Sep 18, 2025):

I initially thought I would have to develop the Modelfile by translating the chat template provided on Hugging Face for qwen3, but I’ve now discovered that each model already has its Modelfile implemented. I really appreciate your help—it was very helpful!

<!-- gh-comment-id:3306955640 --> @redaachouhad commented on GitHub (Sep 18, 2025): I initially thought I would have to develop the Modelfile by translating the chat template provided on Hugging Face for qwen3, but I’ve now discovered that each model already has its Modelfile implemented. I really appreciate your help—it was very helpful!
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#8191
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?