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 #1415] Override SYSTEM parameter by commandline #62791

New Issue
Open
opened 2026-05-03 10:19:26 -05:00 by GiteaMirror · 5 comments
GiteaMirror commented 2026-05-03 10:19:26 -05:00
Owner
Copy Link

Originally created by @marco-trovato on GitHub (Dec 7, 2023).
Original GitHub issue: https://github.com/ollama/ollama/issues/1415

According to the documentation, the only way to change the SYSTEM is to create a new model with modelfile using an existing LLM model already downloaded as template:

ollama create choose-a-model-name -f <location of the file e.g. ./Modelfile>'

But this will copy and duplicate the model file (often bigger than 20 GB)

But using oTerm is possible to change the SYSTEM, please refer to this screenshot for visual reference:
image

REQUEST:
Please add the --system command line to force system, example usage:
ollama run codeup:13b-llama2-chat-q4_K_M --verbose --system "Roleplay as Matrix movie operator before answering the question." "Write Python code to loop for 1 to 10"

Originally created by @marco-trovato on GitHub (Dec 7, 2023). Original GitHub issue: https://github.com/ollama/ollama/issues/1415 According to the [documentation](https://github.com/jmorganca/ollama/blob/main/docs/modelfile.md), the only way to change the SYSTEM is to create a new model with modelfile using an existing LLM model already downloaded as template: `ollama create choose-a-model-name -f <location of the file e.g. ./Modelfile>'` But this will copy and duplicate the model file (often bigger than 20 GB) But using oTerm is possible to change the SYSTEM, please refer to this screenshot for visual reference: ![image](https://github.com/jmorganca/ollama/assets/18162107/42d7b3b6-3e3c-46f1-859d-29755e97ef2a) REQUEST: **Please add the --system command line to force system, example usage:** ollama run codeup:13b-llama2-chat-q4_K_M --verbose **--system** "Roleplay as Matrix movie operator before answering the question." "Write Python code to loop for 1 to 10"
GiteaMirror added the feature request label 2026-05-03 10:19:26 -05:00
GiteaMirror commented 2026-05-03 10:20:08 -05:00
Author
Owner
Copy Link

@BruceMacD commented on GitHub (Dec 7, 2023):

Hi @marco-trovato thanks for opening the issue, the REPL actually supports setting the system parameter as of the most recent release. Here is what that looks like:

ollama run mistral
>>> /set system role play as neo from the matrix
Set system template.

>>> will you take the red pill or the blue pill?

I take the red pill.

Does that meet your needs?

<!-- gh-comment-id:1846037955 --> @BruceMacD commented on GitHub (Dec 7, 2023): Hi @marco-trovato thanks for opening the issue, the REPL actually supports setting the system parameter as of the most recent release. Here is what that looks like: ``` ollama run mistral >>> /set system role play as neo from the matrix Set system template. >>> will you take the red pill or the blue pill? I take the red pill. ``` Does that meet your needs?
GiteaMirror commented 2026-05-03 10:20:11 -05:00
Author
Owner
Copy Link

@mxyng commented on GitHub (Dec 7, 2023):

To clarify, while ollama create will create a new model, the model weights are not duplicated on disk. Layers can be shared by multiple models but only one copy is actually persisted

<!-- gh-comment-id:1846127948 --> @mxyng commented on GitHub (Dec 7, 2023): To clarify, while `ollama create` will create a new model, the model weights are not duplicated on disk. Layers can be shared by multiple models but only one copy is actually persisted
GiteaMirror commented 2026-05-03 10:20:14 -05:00
Author
Owner
Copy Link

@marco-trovato commented on GitHub (Dec 7, 2023):

ollama run mistral

/set system role play as neo from the matrix
Does that meet your needs?

This is very useful and I didn't know it, thank you.

Unfortunately in my specific use case I am trying to use it in non interactive mode as command line for a bash script.
i.e.: ollama run codeup:13b-llama2-chat-q4_K_M --verbose "Write Python code to loop for 1 to 10"

The solution you proposed works only as interactive and requires the user to actually type the command.

<!-- gh-comment-id:1846179682 --> @marco-trovato commented on GitHub (Dec 7, 2023): > ollama run mistral > >>> /set system role play as neo from the matrix > Does that meet your needs? This is very useful and I didn't know it, thank you. Unfortunately in my specific use case I am trying to use it in **non interactive mode** as command line for a bash script. i.e.: `ollama run codeup:13b-llama2-chat-q4_K_M --verbose "Write Python code to loop for 1 to 10"` The solution you proposed works only as interactive and requires the user to actually type the command.
GiteaMirror commented 2026-05-03 10:20:16 -05:00
Author
Owner
Copy Link

@iplayfast commented on GitHub (Dec 15, 2023):

I agree that having all the possible / command internally in ollama would be useful to have as external parameters when starting the program.

<!-- gh-comment-id:1858361592 --> @iplayfast commented on GitHub (Dec 15, 2023): I agree that having all the possible / command internally in ollama would be useful to have as external parameters when starting the program.
GiteaMirror commented 2026-05-03 10:20:17 -05:00
Author
Owner
Copy Link

@cirosantilli commented on GitHub (Mar 18, 2025):

Someone semi-asking to set parameters from CLI: https://github.com/ollama/ollama/issues/2505#issuecomment-2435395442

<!-- gh-comment-id:2733205599 --> @cirosantilli commented on GitHub (Mar 18, 2025): Someone semi-asking to set parameters from CLI: https://github.com/ollama/ollama/issues/2505#issuecomment-2435395442
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 feature request
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#62791
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?