Changing libraries while working on a project syncs heavy dependencies

[jbohnslav]

Hi, I'm looking for advice on how to use UV in my workflow. Right now, changing any code in local library requires 7GB of changes in my Docker image.

My project looks like this

libs
    my_lib/
       # requires pytorch, cuda, etc. 7GB
projects
    my_project/
         # requires my_lib

Docker section, as per the UV docs

RUN mkdir /app 
COPY libs /app/libs 

WORKDIR /app/projects/my_project 
ENV UV_LINK_MODE=copy \ 
    UV_COMPILE_BYTECODE=1 
RUN --mount=type=cache,target=/root/.cache/uv \ 
    --mount=type=bind,source=projects/caption/.python-version,target=.python-version \ 
    --mount=type=bind,source=projects/caption/uv.lock,target=uv.lock \ 
    --mount=type=bind,source=projects/caption/pyproject.toml,target=pyproject.toml \ 
    uv sync --frozen --no-install-project --no-editable 

COPY projects/my_project /app/projects/my_project 
# Sync the project 
RUN --mount=type=cache,target=/root/.cache/uv \ 
    uv sync --frozen --no-editable

And here's the chunk of the pyproject.toml:

dependencies = [ "my_lib",  "ruff>=0.6.4", "setuptools>=75.1.0", ] 

[tool.uv.sources] 
my_lib = { path = "../../libs/my_lib", editable = true }

My issue right now is that when I change library code and rebuild my docker container, changing library code is a change to the docker layer that also installs the large dependencies, so I have to build and push the entire thing. This is pushing like 7GB when I change a single line of text in a local library.

How do I fix it?