Create a Docker container for Machine Learning - docker

Question

I'm not a Docker expert and I'm trying to create a container for machine learning projects. This is mostly for academic purpose, as I'm studying machine learning.
I wrote a dockerfile (and a devcontainer.json file to open the container in vscode) that runs fine until I add the lines to build tensorflow. I found three problems but I don't know what I'm missing:
I got a warning about the fact that Bazel installation isn't a release version
When the building phase arrives at ./configure I can't configure through prompt question
I got an error at the very building phase with
ERROR: The project you're trying to build requires Bazel 5.3.0 (specified in /docklearning/tensorflow/.bazelversion), but it wasn't found in /usr/bin.
This is the Dockerfile:
FROM nvidia/cuda:12.0.0-runtime-ubuntu22.04 as base
ARG USER_UID=1000
#switch to non-interactive frontend
ENV DEBIAN_FRONTEND=noninteractive
WORKDIR /docklearning
ADD . /docklearning
# Install packages
RUN apt-get update -q && apt-get install -q -y --no-install-recommends \
apt-transport-https curl gnupg apt-utils wget gcc g++ npm unzip build-essential ca-certificates curl git gh \
make nano iproute2 nano openssh-client openssl procps \
software-properties-common bzip2 subversion neofetch \
fontconfig && \
curl -fsSL https://bazel.build/bazel-release.pub.gpg | gpg --dearmor >bazel-archive-keyring.gpg && \
mv bazel-archive-keyring.gpg /usr/share/keyrings && \
echo "deb [arch=amd64 signed-by=/usr/share/keyrings/bazel-archive-keyring.gpg] https://storage.googleapis.com/bazel-apt stable jdk1.8" \
| tee /etc/apt/sources.list.d/bazel.list && \
apt update && apt install -q -y bazel && \
apt-get full-upgrade -q -y && \
cd ~ && \
wget https://github.com/ryanoasis/nerd-fonts/releases/download/v2.1.0/Meslo.zip && \
mkdir -p .local/share/fonts && \
unzip Meslo.zip -d .local/share/fonts && \
cd .local/share/fonts && rm *Windows* && \
cd ~ && \
rm Meslo.zip && \
fc-cache -fv && \
apt-get install -y zsh zsh-doc chroma
# Install anaconda
RUN wget https://repo.anaconda.com/archive/Anaconda3-2022.10-Linux-x86_64.sh -O Anaconda.sh && \
/bin/bash Anaconda.sh -b -p /opt/conda && \
rm Anaconda.sh
# Install LSD for ls substitute and clean up
RUN wget https://github.com/Peltoche/lsd/releases/download/0.23.1/lsd_0.23.1_amd64.deb -P /tmp && \
dpkg -i /tmp/lsd_0.23.1_amd64.deb && \
rm /tmp/lsd_0.23.1_amd64.deb && \
apt-get autoremove -y && \
apt-get autoclean && \
rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*/apt/lists/* && \
useradd -r -m -s /bin/bash -u ${USER_UID} docklearning
# Add conda to PATH
ENV PATH=/opt/conda/bin:$PATH
ENV HOME=/home/docklearning
# Make zsh default shell
RUN chsh -s /usr/bin/zsh docklearning
# link conda to /etc/profile.d
RUN ln -s /opt/conda/etc/profile.d/conda.sh /etc/profile.d/conda.sh
# Update Conda Base env
RUN conda update -n base --all -y && \
conda install -c conda-forge bandit opt_einsum keras-preprocessing && \
conda clean -a -q -y
# Download Tensorflow from github and build from source
RUN git clone https://github.com/tensorflow/tensorflow.git && \
cd tensorflow && \
./configure && \
bazel build --config=opt --config=cuda --cxxopt="-mavx2" //tensorflow/tools/pip_package:build_pip_package && \
./bazel-bin/tensorflow/tools/pip_package/build_pip_package /tmp/tensorflow_pkg && \
pip install /tmp/tensorflow_pkg/tensorflow-*.whl && \
cd .. && \
rm -rf tensorflow
# Create shell history file
RUN mkdir ${HOME}/zsh_history && \
chown docklearning ${HOME}/zsh_history && \
mkdir ${HOME}/.ssh
# Switch to internal user
USER docklearning
WORKDIR ${HOME}
# Copy user configuration files
COPY --chown=docklearning ./config/.aliases.sh ./
COPY --chown=docklearning ./config/.bashrc ./
COPY --chown=docklearning ./config/.nanorc ./
# Configure Zsh for internal user
ENV ZSH=${HOME}/.oh-my-zsh
ENV ZSH_CUSTOM=${ZSH}/custom
ENV ZSH_PLUGINS=${ZSH_CUSTOM}/plugins
ENV ZSH_THEMES=${ZSH_CUSTOM}/themes
RUN wget -qO- https://raw.githubusercontent.com/ohmyzsh/ohmyzsh/master/tools/install.sh | zsh || true
RUN git clone --single-branch --branch 'master' --depth 1 https://github.com/zsh-users/zsh-syntax-highlighting.git ${ZSH_PLUGINS}/zsh-syntax-highlighting \
&& git clone --single-branch --branch 'master' --depth 1 https://github.com/zsh-users/zsh-autosuggestions ${ZSH_PLUGINS}/zsh-autosuggestions \
&& git clone --single-branch --depth 1 https://github.com/romkatv/powerlevel10k.git ${ZSH_THEMES}/powerlevel10k
COPY --chown=docklearning ./config/.p10k.zsh ./
COPY --chown=docklearning ./config/.zshrc ./
CMD [ "/bin/zsh" ]

Answer 1

I wouldn't say this question about docker or TensorFlow. It is about package installation and OS configuration.
In this case, you need to fix the bazel version first by specifying it explicitly:
apt install -q -y bazel-5.3.0
But that package is quite wired and doesn't create a symlink, so you need to create it:
ln -s /bin/bazel-5.3.0 /bin/bazel
You can validate if bazel installed by running it:
`bazel --version
bazel 5.3.0`