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 #10800] When using gemma3:4b, the reply sometimes cost a lot of time and the answer it wired #32852

New Issue
Closed
opened 2026-04-22 14:43:50 -05:00 by GiteaMirror · 1 comment
GiteaMirror commented 2026-04-22 14:43:50 -05:00
Owner
Copy Link

Originally created by @ywythu on GitHub (May 21, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10800

What is the issue?

Image

the output is something like this, it costs 1479s for a single question "describe the image", and the answer is repeated a lot of time. What should I do, it there any env param I can add to limit the run time of the task , like if the task cost more than 10s, it willed be killed
by the way, this docker is on the aws L4 server, I deployed 3 ollama with gemma3:4b on it , the vram is 15G, it worked well until the task above append, gpu turned to 100% and 3 ollamas don't work at all

Relevant log output

OS

Linux

GPU

Nvidia

CPU

AMD

Ollama version

0.6.8

Originally created by @ywythu on GitHub (May 21, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10800 ### What is the issue? ![Image](https://github.com/user-attachments/assets/eee88b89-be10-417d-aa28-00c66f52b91c) the output is something like this, it costs 1479s for a single question "describe the image", and the answer is repeated a lot of time. What should I do, it there any env param I can add to limit the run time of the task , like if the task cost more than 10s, it willed be killed by the way, this docker is on the aws L4 server, I deployed 3 ollama with gemma3:4b on it , the vram is 15G, it worked well until the task above append, gpu turned to 100% and 3 ollamas don't work at all ### Relevant log output ```shell ``` ### OS Linux ### GPU Nvidia ### CPU AMD ### Ollama version 0.6.8
GiteaMirror added the bug label 2026-04-22 14:43:50 -05:00
GiteaMirror commented 2026-04-22 14:43:52 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (May 21, 2025):

Set num_predict in the API call, it will stop generating tokens when it reaches the limit.

I deployed 3 ollama with gemma3:4b

Do you mean you run 3 ollama servers, each with a copy of gemma3:4b? Why not set OLLAMA_NUM_PARALLEL=3 and just use one server?

<!-- gh-comment-id:2897985577 --> @rick-github commented on GitHub (May 21, 2025): Set [`num_predict`](https://github.com/ollama/ollama/blob/main/docs/api.md#request-8:~:text=seed%22%3A%2042%2C%0A%20%20%20%20%22-,num_predict,-%22%3A%20100%2C%0A%20%20%20%20%22top_k) in the API call, it will stop generating tokens when it reaches the limit. > I deployed 3 ollama with gemma3:4b Do you mean you run 3 ollama servers, each with a copy of gemma3:4b? Why not set `OLLAMA_NUM_PARALLEL=3` and just use one server?
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#32852
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?