Dependency conflict: label-studio-ml requires pillow~=10.3 but label-studio-sdk requires Pillow>=11.3.0

[Johnwz123]

I’m encountering a dependency conflict when installing with pip, the latest label-studio-ml. Here’s the error:

ERROR: Cannot install -r requirements-base.txt (line 2) and label-studio-ml because these package versions have conflicting dependencies.

The conflict is caused by:
    label-studio-ml 2.0.1.dev0 depends on pillow~=10.3
    label-studio-sdk 1.0.19 depends on Pillow>=11.3.0

To fix this you could try to:
1. loosen the range of package versions you've specified
2. remove package versions to allow pip to attempt to solve the dependency conflict

ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts

[OldEntity]

使用 pip 安装最新的 label-studio-ml 时，遇到了依赖项冲突。错误如下：

ERROR: Cannot install -r requirements-base.txt (line 2) and label-studio-ml because these package versions have conflicting dependencies.

The conflict is caused by:
    label-studio-ml 2.0.1.dev0 depends on pillow~=10.3
    label-studio-sdk 1.0.19 depends on Pillow>=11.3.0

To fix this you could try to:
1. loosen the range of package versions you've specified
2. remove package versions to allow pip to attempt to solve the dependency conflict

ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/topics/dependency-resolution/#dealing-with-dependency-conflicts

I met it too

[askxiaozhang]

Me too.

[drdauntless]

Same here, any resolution would be appreciated

[mac799l]

I am also having this issue. I tried to play around with the versions and commits but kept getting the same error.

[kindkev]

ERROR: Cannot install -r requirements-base.txt (line 2) and label-studio-ml because these package versions have conflicting dependencies.

The conflict is caused by:
label-studio-ml 2.0.1.dev0 depends on pillow~=10.3
label-studio-sdk 1.0.19 depends on Pillow>=11.3.0

To fix this you could try to:

1. loosen the range of package versions you've specified
2. remove package versions to allow pip to attempt to solve the dependency conflict

Yep, still looking for a solution.

[o9ami]

Same here

[jclemmons92]

For those looking for quick resolution and are okay with git cloning the repo, you can git clone the repo and remove the version requirement for pillow in the requirements.txt. Then then run pip install -e /path/to/label-studio-ml-backend

[askxiaozhang]

I’ve resolved the issue. The error occurred when I ran docker-compose up in the label_studio_ml/examples/segment_anything_model/ directory. My solution was to modify some configurations in the Dockerfile within that folder and also update the contents of requirements.txt. After these changes, the service successfully started and inference worked properly in Label Studio.

You can refer to the modified repository here:
https://github.com/askxiaozhang/label-studio-ml-backend.git
Branch: fix/dependency-conflict

[tintin-newb]

I was facing the issue while trying to deploy yolo ML backend. I solved it by changing the requirements-base.txt to the following.

gunicorn==23.0.0
label-studio-ml
label-studio-sdk
rq

I added the rq because I was getting a "ModuleNotFound" error once the build was complete and the app was deployed. This results in a successful build.

However, the app fails to run. Here's the error I am getting when I run docker compose up --build from the "yolo" directory.

$ docker compose up --build
WARN[0000] The "TEST_ENV" variable is not set. Defaulting to a blank string. 
WARN[0000] The "LABEL_STUDIO_API_KEY" variable is not set. Defaulting to a blank string. 
WARN[0000] /home/tintin/Documents/video_analytics/label-studio-ml-backend/label_studio_ml/examples/yolo/docker-compose.yml: the attribute `version` is obsolete, it will be ignored, please remove it to avoid potential confusion 
Compose can now delegate builds to bake for better performance.
 To do so, set COMPOSE_BAKE=true.
[+] Building 60.2s (19/19) FINISHED                                                                                                                            docker:default
 => [yolo internal] load build definition from Dockerfile                                                                                                                0.0s
 => => transferring dockerfile: 1.98kB                                                                                                                                   0.0s
 => [yolo internal] load metadata for docker.io/pytorch/pytorch:2.1.2-cuda12.1-cudnn8-runtime                                                                            1.0s
 => [yolo internal] load .dockerignore                                                                                                                                   0.0s
 => => transferring context: 476B                                                                                                                                        0.0s
 => [yolo stage-0  1/13] FROM docker.io/pytorch/pytorch:2.1.2-cuda12.1-cudnn8-runtime@sha256:3387e598cb94fc248d82e712a65b10931a990cea3a2e76362ca30d135f565de4            0.0s
 => [yolo internal] load build context                                                                                                                                   0.0s
 => => transferring context: 2.08kB                                                                                                                                      0.0s
 => CACHED [yolo stage-0  2/13] WORKDIR /app                                                                                                                             0.0s
 => CACHED [yolo stage-0  3/13] RUN conda update conda -y                                                                                                                0.0s
 => CACHED [yolo stage-0  4/13] RUN --mount=type=cache,target="/var/cache/apt",sharing=locked     --mount=type=cache,target="/var/lib/apt/lists",sharing=locked     apt  0.0s
 => [yolo stage-0  5/13] COPY requirements-base.txt .                                                                                                                    0.0s 
 => [yolo stage-0  6/13] RUN --mount=type=cache,target=/.cache,sharing=locked     pip install -r requirements-base.txt                                                  17.0s 
 => [yolo stage-0  7/13] COPY requirements.txt .                                                                                                                         0.0s 
 => [yolo stage-0  8/13] RUN --mount=type=cache,target=/.cache,sharing=locked     pip3 install -r requirements.txt                                                      11.4s 
 => [yolo stage-0  9/13] COPY requirements-test.txt .                                                                                                                    0.0s 
 => [yolo stage-0 10/13] RUN --mount=type=cache,target=/.cache,sharing=locked     if [ "" = "true" ]; then       pip3 install -r requirements-test.txt;     fi           0.1s 
 => [yolo stage-0 11/13] WORKDIR /app                                                                                                                                    0.0s
 => [yolo stage-0 12/13] COPY . ./                                                                                                                                       0.0s
 => [yolo stage-0 13/13] RUN /bin/sh -c 'if [ ! -f /app/models/yolov8m.pt ]; then     yolo predict model=/app/models/yolov8m.pt source=/app/tests/car.jpg     && yolo   28.6s
 => [yolo] exporting to image                                                                                                                                            2.0s
 => => exporting layers                                                                                                                                                  2.0s
 => => writing image sha256:0cd045d629210e26b7e34e2f9f6e8987ad37468a5a29e44abe4960c2abd97e15                                                                             0.0s
 => => naming to docker.io/humansignal/yolo:v0                                                                                                                           0.0s
 => [yolo] resolving provenance for metadata file                                                                                                                        0.0s
[+] Running 2/2
 ✔ yolo            Built                                                                                                                                                 0.0s 
 ✔ Container yolo  Recreated                                                                                                                                             0.0s 
Attaching to yolo
yolo  | [2025-07-11 09:29:58 +0000] [1] [INFO] Starting gunicorn 23.0.0
yolo  | [2025-07-11 09:29:58 +0000] [1] [INFO] Listening at: http://0.0.0.0:9090 (1)
yolo  | [2025-07-11 09:29:58 +0000] [1] [INFO] Using worker: gthread
yolo  | [2025-07-11 09:29:58 +0000] [7] [INFO] Booting worker with pid: 7
yolo  | [2025-07-11 09:29:59 +0000] [7] [ERROR] Exception in worker process
yolo  | Traceback (most recent call last):
yolo  |   File "/opt/conda/lib/python3.10/site-packages/gunicorn/arbiter.py", line 608, in spawn_worker
yolo  |     worker.init_process()
yolo  |   File "/opt/conda/lib/python3.10/site-packages/gunicorn/workers/gthread.py", line 94, in init_process
yolo  |     super().init_process()
yolo  |   File "/opt/conda/lib/python3.10/site-packages/gunicorn/workers/base.py", line 135, in init_process
yolo  |     self.load_wsgi()
yolo  |   File "/opt/conda/lib/python3.10/site-packages/gunicorn/workers/base.py", line 147, in load_wsgi
yolo  |     self.wsgi = self.app.wsgi()
yolo  |   File "/opt/conda/lib/python3.10/site-packages/gunicorn/app/base.py", line 66, in wsgi
yolo  |     self.callable = self.load()
yolo  |   File "/opt/conda/lib/python3.10/site-packages/gunicorn/app/wsgiapp.py", line 57, in load
yolo  |     return self.load_wsgiapp()
yolo  |   File "/opt/conda/lib/python3.10/site-packages/gunicorn/app/wsgiapp.py", line 47, in load_wsgiapp
yolo  |     return util.import_app(self.app_uri)
yolo  |   File "/opt/conda/lib/python3.10/site-packages/gunicorn/util.py", line 370, in import_app
yolo  |     mod = importlib.import_module(module)
yolo  |   File "/opt/conda/lib/python3.10/importlib/__init__.py", line 126, in import_module
yolo  |     return _bootstrap._gcd_import(name[level:], package, level)
yolo  |   File "<frozen importlib._bootstrap>", line 1050, in _gcd_import
yolo  |   File "<frozen importlib._bootstrap>", line 1027, in _find_and_load
yolo  |   File "<frozen importlib._bootstrap>", line 1006, in _find_and_load_unlocked
yolo  |   File "<frozen importlib._bootstrap>", line 688, in _load_unlocked
yolo  |   File "<frozen importlib._bootstrap_external>", line 883, in exec_module
yolo  |   File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
yolo  |   File "/app/_wsgi.py", line 36, in <module>
yolo  |     from model import YOLO
yolo  |   File "/app/model.py", line 5, in <module>
yolo  |     from label_studio_ml.response import ModelResponse
yolo  | ModuleNotFoundError: No module named 'label_studio_ml.response'
yolo  | [2025-07-11 09:29:59 +0000] [7] [INFO] Worker exiting (pid: 7)
yolo  | [2025-07-11 09:29:59 +0000] [1] [ERROR] Worker (pid:7) exited with code 3
yolo  | [2025-07-11 09:29:59 +0000] [1] [ERROR] Shutting down: Master
yolo  | [2025-07-11 09:29:59 +0000] [1] [ERROR] Reason: Worker failed to boot.
yolo exited with code 3

[heidi-humansignal]

Hello,

The issue has been fixed. Could you please pull the latest master and run label Studio again?
Thank you and apologies for any inconvenience this might have caused.

Best Regards!

Comment by Oussama Assili
Workflow Run