I am trying to create a docker image that has a conda environment with PyTorch, but I get an error message source: not found
.
Dockerfile
FROM continuumio/miniconda3
FROM anibali/pytorch:1.8.1-cuda11.1-ubuntu20.04
USER user
COPY --chown=user ./environment.yaml /home/user/app/environment.yaml
RUN conda init bash && source /home/user/.bashrc && conda activate my_env
WORKDIR /home/user/app
but I get this error message:
Step 5/6 : RUN conda init bash && source /home/user/.bashrc && conda activate my_env
---> [Warning] Your kernel does not support swap limit capabilities or the cgroup is not mounted. Memory limited without swap.
---> Running in f435bbf4b511
no change /home/user/miniconda/condabin/conda
no change /home/user/miniconda/bin/conda
no change /home/user/miniconda/bin/conda-env
no change /home/user/miniconda/bin/activate
no change /home/user/miniconda/bin/deactivate
no change /home/user/miniconda/etc/profile.d/conda.sh
no change /home/user/miniconda/etc/fish/conf.d/conda.fish
no change /home/user/miniconda/shell/condabin/Conda.psm1
no change /home/user/miniconda/shell/condabin/conda-hook.ps1
no change /home/user/miniconda/lib/python3.8/site-packages/xontrib/conda.xsh
no change /home/user/miniconda/etc/profile.d/conda.csh
modified /home/user/.bashrc
==> For changes to take effect, close and re-open your current shell. <==
/bin/sh: 1: source: not found
The command '/bin/sh -c conda init bash && source /home/user/.bashrc && conda activate my_env' returned a non-zero code: 127
How can I solve this? I know how to use pip. but I need conda to have best reproducibility with one cluster that doesn’t run docker.