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 #14016] Docker 0.15.3 failed to initialize MLX #71221

New Issue
Open
opened 2026-05-05 00:43:17 -05:00 by GiteaMirror · 12 comments
GiteaMirror commented 2026-05-05 00:43:17 -05:00
Owner
Copy Link

Originally created by @Slawka on GitHub (Feb 1, 2026).
Original GitHub issue: https://github.com/ollama/ollama/issues/14016

What is the issue?

image: ollama/ollama:0.15.3

The problem occurred in version 0.15.3
ollama run x/z-image-turbo:latest
ollama run x/flux2-klein

Relevant log output

root@907ff5fce378:/# ollama run x/z-image-turbo:latest
Error: failed to load model: 500 Internal Server Error: image runner failed: Error: failed to initialize MLX: MLX: Failed to load libmlxc library. Tried: ./build/lib/ollama/libmlxc.so, libmlxc.so. Last error: libquadmath.so.0: cannot open shared object file: No such file or directory (exit: exit status 1)

OS

Docker

GPU

Nvidia

CPU

Intel

Ollama version

0.15.3

Originally created by @Slawka on GitHub (Feb 1, 2026). Original GitHub issue: https://github.com/ollama/ollama/issues/14016 ### What is the issue? image: ollama/ollama:0.15.3 The problem occurred in version 0.15.3 ollama run x/z-image-turbo:latest ollama run x/flux2-klein ### Relevant log output ```shell root@907ff5fce378:/# ollama run x/z-image-turbo:latest Error: failed to load model: 500 Internal Server Error: image runner failed: Error: failed to initialize MLX: MLX: Failed to load libmlxc library. Tried: ./build/lib/ollama/libmlxc.so, libmlxc.so. Last error: libquadmath.so.0: cannot open shared object file: No such file or directory (exit: exit status 1) ``` ### OS Docker ### GPU Nvidia ### CPU Intel ### Ollama version 0.15.3
GiteaMirror added the bug label 2026-05-05 00:43:17 -05:00
GiteaMirror commented 2026-05-05 00:43:21 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 1, 2026):

https://github.com/ollama/ollama/pull/13984

<!-- gh-comment-id:3831729243 --> @rick-github commented on GitHub (Feb 1, 2026): https://github.com/ollama/ollama/pull/13984
GiteaMirror commented 2026-05-05 00:43:22 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 1, 2026):

Use the following Dockerfile to create an ollama container with the required installs without doing a full build:

FROM ollama/ollama:latest
RUN apt update && apt install -y libquadmath0 wget
ENV CUDA_LIBS="                                           \
        cuda-toolkit-config-common_13.0.48-1_all.deb      \
        cuda-toolkit-13-config-common_13.0.48-1_all.deb   \
        cuda-toolkit-13-0-config-common_13.0.48-1_all.deb \
        cuda-cudart-13-0_13.0.48-1_amd64.deb              \
        cuda-cccl-13-0_13.0.50-1_amd64.deb                \
        cuda-culibos-dev-13-0_13.0.39-1_amd64.deb         \
        cuda-driver-dev-13-0_13.0.48-1_amd64.deb          \
        cuda-cudart-dev-13-0_13.0.48-1_amd64.deb          \
        cuda-nvrtc-13-0_13.0.48-1_amd64.deb               \
        "
RUN for i in $CUDA_LIBS ; do wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/$i ; done && dpkg -i $CUDA_LIBS
RUN cd /usr/local/cuda-13.0/targets/x86_64-linux && ln -s cccl/cuda include/cuda && ln -s /usr/local/cuda-13.0/targets/x86_64-linux/include /usr/lib/ollama/mlx_cuda_v13/include

$ docker build -t ollama/ollama:imagegen . -f Dockerfile
$ docker run --rm -d --gpus all -v /usr/share/ollama/.ollama:/root/.ollama --name imagegen ollama/ollama:imagegen
$ docker exec -it imagegen ollama run x/z-image-turbo:fp8 "tux penguin and llama sitting side by side, with a sign between them that says 'docker'"
Image
<!-- gh-comment-id:3831904450 --> @rick-github commented on GitHub (Feb 1, 2026): Use the following Dockerfile to create an ollama container with the required installs without doing a full build: ```dockerfile FROM ollama/ollama:latest RUN apt update && apt install -y libquadmath0 wget ENV CUDA_LIBS=" \ cuda-toolkit-config-common_13.0.48-1_all.deb \ cuda-toolkit-13-config-common_13.0.48-1_all.deb \ cuda-toolkit-13-0-config-common_13.0.48-1_all.deb \ cuda-cudart-13-0_13.0.48-1_amd64.deb \ cuda-cccl-13-0_13.0.50-1_amd64.deb \ cuda-culibos-dev-13-0_13.0.39-1_amd64.deb \ cuda-driver-dev-13-0_13.0.48-1_amd64.deb \ cuda-cudart-dev-13-0_13.0.48-1_amd64.deb \ cuda-nvrtc-13-0_13.0.48-1_amd64.deb \ " RUN for i in $CUDA_LIBS ; do wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/$i ; done && dpkg -i $CUDA_LIBS RUN cd /usr/local/cuda-13.0/targets/x86_64-linux && ln -s cccl/cuda include/cuda && ln -s /usr/local/cuda-13.0/targets/x86_64-linux/include /usr/lib/ollama/mlx_cuda_v13/include ``` ```console $ docker build -t ollama/ollama:imagegen . -f Dockerfile $ docker run --rm -d --gpus all -v /usr/share/ollama/.ollama:/root/.ollama --name imagegen ollama/ollama:imagegen $ docker exec -it imagegen ollama run x/z-image-turbo:fp8 "tux penguin and llama sitting side by side, with a sign between them that says 'docker'" ``` <img width="993" height="1010" alt="Image" src="https://github.com/user-attachments/assets/54b87f1f-d05d-49e9-9f9f-7d90491f5f87" />
GiteaMirror commented 2026-05-05 00:43:24 -05:00
Author
Owner
Copy Link

@Slawka commented on GitHub (Feb 8, 2026):

ollama run x/z-image-turbo:fp8 "tux penguin and llama sitting side by side, with a sign between them that says 'docker'"

`root@f57de88e0670:/# ollama run x/flux2-klein "tux penguin and llama sitting side by side, with a sign between them that says 'docker'"

Error: 500 Internal Server Error: mlx runner exited unexpectedly: exit status 255"
`

<!-- gh-comment-id:3867527920 --> @Slawka commented on GitHub (Feb 8, 2026): > ollama run x/z-image-turbo:fp8 "tux penguin and llama sitting side by side, with a sign between them that says 'docker'" `root@f57de88e0670:/# ollama run x/flux2-klein "tux penguin and llama sitting side by side, with a sign between them that says 'docker'" Error: 500 Internal Server Error: mlx runner exited unexpectedly: exit status 255" `
GiteaMirror commented 2026-05-05 00:43:26 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 8, 2026):

Logs?

<!-- gh-comment-id:3867541143 --> @rick-github commented on GitHub (Feb 8, 2026): Logs?
GiteaMirror commented 2026-05-05 00:43:28 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 8, 2026):

Probably #14046.

<!-- gh-comment-id:3867543555 --> @rick-github commented on GitHub (Feb 8, 2026): Probably #14046.
GiteaMirror commented 2026-05-05 00:43:31 -05:00
Author
Owner
Copy Link

@Hello-World-Traveler commented on GitHub (Feb 14, 2026):

@rick-github Do you know when your PR for docker will be added?

<!-- gh-comment-id:3901725648 --> @Hello-World-Traveler commented on GitHub (Feb 14, 2026): @rick-github Do you know when your PR for docker will be added?
GiteaMirror commented 2026-05-05 00:43:32 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 14, 2026):

No idea. I have PRs that are over a year old, I have no clue what conditions have to be met for a PR to even be reviewed let alone approved.

<!-- gh-comment-id:3902162191 --> @rick-github commented on GitHub (Feb 14, 2026): No idea. I have PRs that are over a year old, I have no clue what conditions have to be met for a PR to even be reviewed let alone approved.
GiteaMirror commented 2026-05-05 00:43:34 -05:00
Author
Owner
Copy Link

@Hello-World-Traveler commented on GitHub (Feb 14, 2026):

@rick-github Thank you for the PR. I did notice that it hasn't been requested for a review yet, could that also be why the delay?

<!-- gh-comment-id:3902458639 --> @Hello-World-Traveler commented on GitHub (Feb 14, 2026): @rick-github Thank you for the PR. I did notice that it hasn't been requested for a review yet, could that also be why the delay?
GiteaMirror commented 2026-05-05 00:43:35 -05:00
Author
Owner
Copy Link

@onoraba commented on GitHub (Mar 19, 2026):

$ ollama run x/z-image-turbo:bf16 "test"

Error: 500 Internal Server Error: mlx runner failed: Error: failed to initialize MLX: libmlxc.so not found (exit: exit status 1)

alpine v3.23.3, ollama v0.18.0

<!-- gh-comment-id:4092862063 --> @onoraba commented on GitHub (Mar 19, 2026): $ ollama run x/z-image-turbo:bf16 "test" Error: 500 Internal Server Error: mlx runner failed: Error: failed to initialize MLX: libmlxc.so not found (exit: exit status 1) alpine v3.23.3, ollama v0.18.0
GiteaMirror commented 2026-05-05 00:43:37 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Mar 19, 2026):

How was ollama installed?

<!-- gh-comment-id:4092895757 --> @rick-github commented on GitHub (Mar 19, 2026): How was ollama installed?
GiteaMirror commented 2026-05-05 00:43:39 -05:00
Author
Owner
Copy Link

@adjenks commented on GitHub (Apr 21, 2026):

I'm getting:

"failed to initialize MLX: MLX: Failed to load /usr/lib/ollama/mlx_cuda_v13/libmlxc.so"

in the latest docker container.

<!-- gh-comment-id:4287382485 --> @adjenks commented on GitHub (Apr 21, 2026): I'm getting: "failed to initialize MLX: MLX: Failed to load /usr/lib/ollama/mlx_cuda_v13/libmlxc.so" in the latest docker container.
GiteaMirror commented 2026-05-05 00:43:40 -05:00
Author
Owner
Copy Link

@jeff-kelley commented on GitHub (Apr 22, 2026):

On bare metal :

$ ollama run x/z-image-turbo
Error: failed to load model: 500 Internal Server Error: mlx runner failed: Error: failed to initialize MLX: MLX: Failed to load /usr/local/lib/ollama/mlx_cuda_v13/libmlxc.so: libmlx.so: cannot open shared object file: No such file or directory (exit: exit status 1)

$ ls -l /usr/local/lib/ollama/mlx_cuda_v13/libmlxc.so
-rwxr-xr-x 1 root root 674360 Apr 22 04:22 /usr/local/lib/ollama/mlx_cuda_v13/libmlxc.so

ollama version: 0.21.1
NVIDIA Driver Version: 550.163.01
CUDA Version: 12.4
Debian 13

<!-- gh-comment-id:4300653634 --> @jeff-kelley commented on GitHub (Apr 22, 2026): On bare metal : ``` $ ollama run x/z-image-turbo Error: failed to load model: 500 Internal Server Error: mlx runner failed: Error: failed to initialize MLX: MLX: Failed to load /usr/local/lib/ollama/mlx_cuda_v13/libmlxc.so: libmlx.so: cannot open shared object file: No such file or directory (exit: exit status 1) $ ls -l /usr/local/lib/ollama/mlx_cuda_v13/libmlxc.so -rwxr-xr-x 1 root root 674360 Apr 22 04:22 /usr/local/lib/ollama/mlx_cuda_v13/libmlxc.so ``` ollama version: 0.21.1 NVIDIA Driver Version: 550.163.01 CUDA Version: 12.4 Debian 13
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#71221
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?