github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-05 23:53:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #7716] Feature suggestions and development compilation environment issues #30686

New Issue
Open
opened 2026-04-22 10:35:17 -05:00 by GiteaMirror · 3 comments
GiteaMirror commented 2026-04-22 10:35:17 -05:00
Owner
Copy Link

Originally created by @mingyue0094 on GitHub (Nov 18, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/7716

Wish:

  1. set env avx=0 will automatically try to use Nvidia gpu
  2. On this repository page, press . to enter a complete development environment to modify code, compile, download files, and run tests. Configuring this development environment is so complicated and difficult.

Good luck to you

set env OLLAMA_HOST "0.0.0.0" can control the listening address. Many home users have old computers, but they can still be used. The motherboard of this old computer does not support AVX, but for gaming, it has NVIDIA GPU, such as 3060.

So, in the future, can we add an environment variable as a switch, with avx = 0, so that the GPU without avx is adopted?

I read the Issues and saw that many people enthusiastically shared how to modify the logic and compile and implement this function by themselves. This is particularly unfriendly for people who are not doing cross-platform development in Go, because they will be stuck in the environment preparation part. Maybe the source code can be changed correctly, but because the environment is very complicated, it cannot be compiled. Moreover, people who are not doing cross-platform development in Go will only compile this environment once and will not need it later.

If your official support can try to support GPU without avx when users modify and add an environment variable, it will be great.
Or, it would be good to have a development compilation environment. Just like, open github, press . and then start modifying the code, perform compilation, download the compiled file and test it.

Originally created by @mingyue0094 on GitHub (Nov 18, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/7716 Wish: 1. set env avx=0 will automatically try to use Nvidia gpu 2. On this repository page, press `.` to enter a complete development environment to modify code, compile, download files, and run tests. Configuring this development environment is so complicated and difficult. Good luck to you ------- set env OLLAMA_HOST "0.0.0.0" can control the listening address. Many home users have old computers, but they can still be used. The motherboard of this old computer does not support AVX, but for gaming, it has NVIDIA GPU, such as 3060. So, in the future, can we add an environment variable as a switch, with `avx = 0`, so that the GPU without avx is adopted? I read the Issues and saw that many people enthusiastically shared how to modify the logic and compile and implement this function by themselves. This is particularly unfriendly for people who are not doing cross-platform development in Go, because they will be stuck in the environment preparation part. Maybe the source code can be changed correctly, but because the environment is very complicated, it cannot be compiled. Moreover, people who are not doing cross-platform development in Go will only compile this environment once and will not need it later. If your official support can try to support GPU without avx when users modify and add an environment variable, it will be great. Or, it would be good to have a development compilation environment. Just like, open github, press `.` and then start modifying the code, perform compilation, download the compiled file and test it.
GiteaMirror added the feature request label 2026-04-22 10:35:17 -05:00
GiteaMirror commented 2026-04-22 10:35:18 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Nov 19, 2024):

Fully dynamic vector feature selection at runtime will require a more extensive change to the architecture.

Once #7499 merges, the intent is to make this relatively straight forward for people to build locally.

<!-- gh-comment-id:2484440735 --> @dhiltgen commented on GitHub (Nov 19, 2024): Fully dynamic vector feature selection at runtime will require a more extensive change to the architecture. Once #7499 merges, the intent is to make this relatively straight forward for people to build locally.
GiteaMirror commented 2026-04-22 10:35:18 -05:00
Author
Owner
Copy Link

@mingyue0094 commented on GitHub (Nov 20, 2024):

@dhiltgen
I see your build workflows

this build env ， There are so many steps. Too difficult

According to your workflows, you should configure the github key and modify some files, then you can automatically build the exe and upload it to github. However, I don't understand which files to modify and how to modify them. It's so difficult.

<!-- gh-comment-id:2489407407 --> @mingyue0094 commented on GitHub (Nov 20, 2024): @dhiltgen I see your build [workflows](https://raw.githubusercontent.com/dhiltgen/ollama/2a0716598ab7128f91012310e7ce44bda5046142/.github/workflows/release.yaml) this build env ， There are so many steps. Too difficult According to your workflows, you should configure the github key and modify some files, then you can automatically build the exe and upload it to github. However, I don't understand which files to modify and how to modify them. It's so difficult.
GiteaMirror commented 2026-04-22 10:35:19 -05:00
Author
Owner
Copy Link

@mingyue0094 commented on GitHub (Nov 20, 2024):

In addition, the source code should also be pulled from your repository

<!-- gh-comment-id:2489409438 --> @mingyue0094 commented on GitHub (Nov 20, 2024): In addition, the source code should also be pulled from your repository
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#30686
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?