github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 13:19:40 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #4784] Weird output with ordinary setting #28773

New Issue
Closed
opened 2026-04-22 07:18:08 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-22 07:18:08 -05:00
Owner
Copy Link

Originally created by @JoonSumisu on GitHub (Jun 2, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4784

What is the issue?

Hello guys

  1. I use the a q4 llama 70b from NousResearch/Meta-Llama-3-70B-GGUF Meta-Llama-3-70B-Q4_K_M.gguf

*I am not really sure this model function well

  1. with correct FROM and

TEMPLATE """{{ if .System }}<|start_header_id|>system<|end_header_id|>

{{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|>

{{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|>

{{ .Response }}<|eot_id|>
"""

  1. the output is weird like

hello
This is a simple program that uses an if-else chain to determine whether someone has low education based on their
years of schooling. The input() function prompts the user for information while the int() function converts the
string into an integer value so it can be used in comparison expressions like x >= y etc.. If no input was given
or not a valid number then None will be returned which causes all subsequent operations to fail due to type
mismatch errors but we still try by assuming

  1. should I give it some correct parameters?

I use another model with follow instruction, the output is "normal"
Any advice where I can find those correct parameters?

  1. And how can I stop loading the model?
    Sorry for the stupid question
    so after I run a model, I have to kill the whole ollama for free the vram
    Available Commands has no command for stop the loading

best

OS

WSL2

GPU

Nvidia

CPU

Intel

Ollama version

0.1.39

Originally created by @JoonSumisu on GitHub (Jun 2, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4784 ### What is the issue? Hello guys 1. I use the a q4 llama 70b from NousResearch/Meta-Llama-3-70B-GGUF Meta-Llama-3-70B-Q4_K_M.gguf *I am not really sure this model function well 3. with correct FROM and TEMPLATE """{{ if .System }}<|start_header_id|>system<|end_header_id|> {{ .System }}<|eot_id|>{{ end }}{{ if .Prompt }}<|start_header_id|>user<|end_header_id|> {{ .Prompt }}<|eot_id|>{{ end }}<|start_header_id|>assistant<|end_header_id|> {{ .Response }}<|eot_id|> """ 4. the output is weird like >>> hello This is a simple program that uses an if-else chain to determine whether someone has low education based on their years of schooling. The input() function prompts the user for information while the int() function converts the string into an integer value so it can be used in comparison expressions like x >= y etc.. If no input was given or not a valid number then None will be returned which causes all subsequent operations to fail due to type mismatch errors but we still try by assuming 5. should I give it some correct parameters? I use another model with follow instruction, the output is "normal" Any advice where I can find those correct parameters? 6. And how can I stop loading the model? Sorry for the stupid question so after I run a model, I have to kill the whole ollama for free the vram Available Commands has no command for stop the loading best ### OS WSL2 ### GPU Nvidia ### CPU Intel ### Ollama version 0.1.39
GiteaMirror added the bug label 2026-04-22 07:18:08 -05:00
GiteaMirror commented 2026-04-22 07:18:09 -05:00
Author
Owner
Copy Link

@pdevine commented on GitHub (Jun 5, 2024):

@JoonSumisu To unload the model from the CLI use ollama run --keep-alive=0 <model> "" I'm not sure what parameters that particular model uses. Usually we find when a model is giving strange output it's because the template is incorrect, although it is possible some of the other parameters are incorrect.

I would check with wherever you got the binary file or w/ Nous to double check if you need to set anything.

<!-- gh-comment-id:2150947882 --> @pdevine commented on GitHub (Jun 5, 2024): @JoonSumisu To unload the model from the CLI use `ollama run --keep-alive=0 <model> ""` I'm not sure what parameters that particular model uses. Usually we find when a model is giving strange output it's because the template is incorrect, although it is possible some of the other parameters are incorrect. I would check with wherever you got the binary file or w/ Nous to double check if you need to set anything.
GiteaMirror commented 2026-04-22 07:18:10 -05:00
Author
Owner
Copy Link

@JoonSumisu commented on GitHub (Jun 6, 2024):

Contributor

Thank you so much pdevine!

with the default template and parameters, the output was also strange. Should I conclude that it is a dysfunctional model?
And if possible , may I suggest a feature that allow the user to unload the model by input /exit or /unload for free the vram after done with the chat, thank you.

<!-- gh-comment-id:2152289506 --> @JoonSumisu commented on GitHub (Jun 6, 2024): > Contributor Thank you so much pdevine! with the default template and parameters, the output was also strange. Should I conclude that it is a dysfunctional model? And if possible , may I suggest a feature that allow the user to unload the model by input /exit or /unload for free the vram after done with the chat, thank you.
GiteaMirror commented 2026-04-22 07:18:12 -05:00
Author
Owner
Copy Link

@JoonSumisu commented on GitHub (Jun 12, 2024):

Hi pdevine

So today, I retest the llama3

with the official ollama run llama3:70b-text-q4_K_M

The output is also bugging.
image

** by update the ollama, the output get usual

<!-- gh-comment-id:2163289600 --> @JoonSumisu commented on GitHub (Jun 12, 2024): Hi pdevine So today, I retest the llama3 with the official ollama run llama3:70b-text-q4_K_M The output is also bugging. <img width="695" alt="image" src="https://github.com/ollama/ollama/assets/77182770/d664a35e-6e1e-49d5-9919-b1a7ddda9d3d"> ** by update the ollama, the output get usual
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#28773
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?