Docker image

[Joschka]

Hi, do you have a pre-built docker image that can be used to run DiGress by any chance? I would highly appreciate that.

[Clement Vignac]

Hello,
we used one a long time ago, but it's probably not up-to-date. You can start from this if you want to build one, but don't expect it to work from the first try.

FROM nvcr.io/nvidia/pytorch:22.05-py3
# update conda, install mamba so everything else will be faster, then install rdkit and other conda dependencies, then install our package
RUN conda update -n base -y conda && conda install -n base -c conda-forge -y mamba #&&
RUN conda install -n base -c conda-forge graph-tool #&&
RUN mamba  install -y rdkit #&& 
RUN pip install pyyaml && pip install overrides imageio numpy scipy tqdm wandb hydra-core seaborn #&&
RUN pip install pytorch_lightning==1.6  torchmetrics torch==1.11 torchvision  --extra-index-url https://download.pytorch.org/whl/cu113 #&&
RUN pip install torch-scatter torch-sparse torch-cluster torch-spline-conv torch-geometric -f https://data.pyg.org/whl/torch-1.11.0+cu113.html
#RUN pip install pyyaml && pip install overrides imageio numpy scipy tqdm wandb hydra-core #&&
ADD src/analysis /workspace/analysis
ADD configs /workspace/configs
ADD src/diffusion /workspace/diffusion
ADD src/models /workspace/models
ADD ./data/qm9/qm9_pyg /workspace/data/qm9/qm9_pyg
ADD *.py /workspace/
RUN pip install -e .
# expects you to to run the image with `docker run -e WANDB_API_KEY=$YOUR_API_KEY graphgendiff:latest`
CMD python main.py

[Joschka]

Thanks!