github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 08:30:05 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #8696] No more binaries? #5635

New Issue
Closed
opened 2026-04-12 16:55:14 -05:00 by GiteaMirror · 4 comments
GiteaMirror commented 2026-04-12 16:55:14 -05:00
Owner
Copy Link

Originally created by @regularRandom on GitHub (Jan 30, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/8696

What is the issue?

After the latest merge there is no ollama binary which can be used in Systemd service. There are only following .so:

  • libggml-base.so
  • libggml-cpu-alderlake.so
  • libggml-cpu-haswell.so
  • libggml-cpu-icelake.so
  • libggml-cpu-sandybridge.so
  • libggml-cpu-sapphirerapids.so
  • libggml-cpu-skylakex.so
  • libggml-cuda.so

I tried with the new build sequence (customized a bit):

cmake --install-prefix /opt/ollama --preset CUDA -B build
cmake --build --preset CUDA build

but only libraries above generated in the build directory, and that's all. Previous approach was something like that:

make -j -S cuda_avx2
go build .

As a result I was getting ./ollama binary and couple of runner's libs. And that was perfectly working as Systemd service. Now it doesn't.

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

0.5.7-7

Originally created by @regularRandom on GitHub (Jan 30, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/8696 ### What is the issue? After the latest merge there is no ollama binary which can be used in Systemd service. There are only following .so: - libggml-base.so - libggml-cpu-alderlake.so - libggml-cpu-haswell.so - libggml-cpu-icelake.so - libggml-cpu-sandybridge.so - libggml-cpu-sapphirerapids.so - libggml-cpu-skylakex.so - libggml-cuda.so I tried with the new build sequence (customized a bit): ``` cmake --install-prefix /opt/ollama --preset CUDA -B build cmake --build --preset CUDA build ``` but only libraries above generated in the build directory, and that's all. Previous approach was something like that: ``` make -j -S cuda_avx2 go build . ``` As a result I was getting ./ollama binary and couple of runner's libs. And that was perfectly working as Systemd service. Now it doesn't. ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version >0.5.7-7
GiteaMirror added the bug label 2026-04-12 16:55:14 -05:00
GiteaMirror commented 2026-04-12 16:55:15 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Jan 30, 2025):

Hi @regularRandom, go build . is still required as a separate step to build the Go binary. And so running:

cmake -B build
cmake --build build
go build .
./ollama serve

Will create the binary. The binary looks for the .so files under ../lib/ollama or build/lib/ollama (for development) relative to ollama

Sorry about that. Let me know if you're still hitting issues.

<!-- gh-comment-id:2624909393 --> @jmorganca commented on GitHub (Jan 30, 2025): Hi @regularRandom, `go build .` is still required as a separate step to build the Go binary. And so running: ``` cmake -B build cmake --build build go build . ./ollama serve ``` Will create the binary. The binary looks for the .so files under `../lib/ollama` or `build/lib/ollama` (for development) relative to `ollama` Sorry about that. Let me know if you're still hitting issues.
GiteaMirror commented 2026-04-12 16:55:16 -05:00
Author
Owner
Copy Link

@regularRandom commented on GitHub (Jan 30, 2025):

Yes, go build . made the trick. Probably it needs to be reflected in the developer guide.

Thank you.

<!-- gh-comment-id:2624922721 --> @regularRandom commented on GitHub (Jan 30, 2025): Yes, **go build .** made the trick. Probably it needs to be reflected in the developer guide. Thank you.
GiteaMirror commented 2026-04-12 16:55:17 -05:00
Author
Owner
Copy Link

@user-33948 commented on GitHub (Feb 1, 2025):

Hi @regularRandom, go build . is still required as a separate step to build the Go binary. And so running:

cmake -B build
cmake --build build
go build .
./ollama serve

Will create the binary. The binary looks for the .so files under ../lib/ollama or build/lib/ollama (for development) relative to ollama

Sorry about that. Let me know if you're still hitting issues.

I had a previous Docker image that relied on Ollama, so I had copy/pasted the ollama binary and the models into my docker image. I went to update the Ollama binary i had (403,976kb), and similarly noticed there is no longer an Ollama binary that gets downloaded when you download Ollama to Linux.

@jmorganca Given this -- what do I need to make this Ollama accessible in a custom Docker image? If there is no longer a large Binary file, what file (s) are needed to make ollama accessible / where do they get sent during the download process?

<!-- gh-comment-id:2629034699 --> @user-33948 commented on GitHub (Feb 1, 2025): > Hi [@regularRandom](https://github.com/regularRandom), `go build .` is still required as a separate step to build the Go binary. And so running: > > ``` > cmake -B build > cmake --build build > go build . > ./ollama serve > ``` > > Will create the binary. The binary looks for the .so files under `../lib/ollama` or `build/lib/ollama` (for development) relative to `ollama` > > Sorry about that. Let me know if you're still hitting issues. I had a previous Docker image that relied on Ollama, so I had copy/pasted the ollama binary and the models into my docker image. I went to update the Ollama binary i had (403,976kb), and similarly noticed there is no longer an Ollama binary that gets downloaded when you download Ollama to Linux. @jmorganca Given this -- what do I need to make this Ollama accessible in a custom Docker image? If there is no longer a large Binary file, what file (s) are needed to make ollama accessible / where do they get sent during the download process?
GiteaMirror commented 2026-04-12 16:55:18 -05:00
Author
Owner
Copy Link

@user-33948 commented on GitHub (Feb 22, 2025):

@jmorganca Wanted to follow up on the question above. Is there still a large (~400kb) Ollama binary that Ollama runs on? In subsequent updates if that has been removed, does it impact what would be required to build a docker image with Ollama?

<!-- gh-comment-id:2676220032 --> @user-33948 commented on GitHub (Feb 22, 2025): @jmorganca Wanted to follow up on the question above. Is there still a large (~400kb) Ollama binary that Ollama runs on? In subsequent updates if that has been removed, does it impact what would be required to build a docker image with Ollama?
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#5635
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?