entrypoint.sh

#!/bin/sh

set -e

# Set up display; otherwise rendering will fail
Xvfb -screen 0 320x240x24 &
export DISPLAY=:0

# Wait for the file to come up
file="/tmp/.X11-unix/X0"
for i in $(seq 1 10); do
    if [ -e "$file" ]; then
	break
    fi

    echo "Waiting for $file to be created (try $i/10)"
    sleep "$i"
done
if ! [ -e "$file" ]; then
    echo "Timing out: $file was not created"
    exit 1
fi

exec "$@"

GPU docker

# Base softlearning container that contains all softlearning requirements,
# but not the actual softlearning repo. Could be used for example when developing
# softlearning, in which case you would mount softlearning repo in to the container
# as a volume, and thus be able to modify code on the host, yet run things inside
# the container. You are encouraged to use docker-compose (docker-compose.dev.yml),
# which should allow you to setup your environment with a single one command.

FROM nvidia/cuda:9.0-runtime-ubuntu16.04

ARG MJKEY

MAINTAINER Kristian Hartikainen <[email protected]>

ENV LANG=C.UTF-8 LC_ALL=C.UTF-8
ENV PATH /opt/conda/bin:$PATH

RUN apt-get update --fix-missing && \
    apt-get install -y wget bzip2 ca-certificates curl git && \
    apt-get clean && \
    rm -rf /var/lib/apt/lists/*

RUN wget --quiet https://repo.anaconda.com/miniconda/Miniconda3-4.5.4-Linux-x86_64.sh -O ~/miniconda.sh && \
    /bin/bash ~/miniconda.sh -b -p /opt/conda && \
    rm ~/miniconda.sh && \
    /opt/conda/bin/conda clean -tipsy && \
    ln -s /opt/conda/etc/profile.d/conda.sh /etc/profile.d/conda.sh && \
    echo ". /opt/conda/etc/profile.d/conda.sh" >> ~/.bashrc && \
    echo "conda activate base" >> ~/.bashrc

RUN conda update -y --name base conda

# ========== Tensorflow dependencies ==========
RUN apt-get update \
    && apt-get upgrade -y \
    && apt-get install -y --no-install-recommends \
        build-essential \
        cuda-command-line-tools-9-0 \
        cuda-cublas-9-0 \
        cuda-cufft-9-0 \
        cuda-curand-9-0 \
        cuda-cusolver-9-0 \
        cuda-cusparse-9-0 \
        curl \
        libcudnn7=7.1.4.18-1+cuda9.0 \
        libnccl2=2.2.13-1+cuda9.0 \
        libfreetype6-dev \
        libhdf5-serial-dev \
        libpng12-dev \
        libzmq3-dev \
        pkg-config \
        python \
        python-dev \
        rsync \
        software-properties-common \
        unzip \
        gcc \
    && apt-get clean \
    && rm -rf /var/lib/apt/lists/*


# ========== Softlearning dependencies ==========
RUN apt-get update \
    # DO NOT apt-get upgrade here, it'll fuck up the tensorflow dependencies
    && apt-get install -y --no-install-recommends \
        build-essential \
        curl \
        git \
        make \
        cmake \
        swig \
        libz-dev \
        unzip \
        zlib1g-dev \
        libglfw3 \
        libglfw3-dev \
        libxrandr2 \
        libxinerama-dev \
        libxi6 \
        libxcursor-dev \
        libgl1-mesa-dev \
        libgl1-mesa-glx \
        libglew-dev \
        libosmesa6-dev \
        ack-grep \
        patchelf \
        vim \
        emacs \
        wget \
        xpra \
        xserver-xorg-dev \
        xvfb \
    && apt-get clean \
    && rm -rf /var/lib/apt/lists/*

# ========= Google Cloud SDK ===========
RUN export CLOUD_SDK_REPO="cloud-sdk-$(lsb_release -c -s)" \
    && echo "deb http://packages.cloud.google.com/apt $CLOUD_SDK_REPO main" \
            | tee -a /etc/apt/sources.list.d/google-cloud-sdk.list \
    && curl https://packages.cloud.google.com/apt/doc/apt-key.gpg \
            | apt-key add - \
    && apt-get update -y \
    && apt-get install google-cloud-sdk -y

# ========= MuJoCo ===============
# Rllab requires mujoco 1.31
ENV MUJOCO_VERSION=131 \
    MUJOCO_PATH=/root/.mujoco

RUN MUJOCO_ZIP="mjpro${MUJOCO_VERSION}_linux.zip" \
    && mkdir -p ${MUJOCO_PATH} \
    && wget -P ${MUJOCO_PATH} https://www.roboti.us/download/${MUJOCO_ZIP} \
    && unzip ${MUJOCO_PATH}/${MUJOCO_ZIP} -d ${MUJOCO_PATH} \
    && rm ${MUJOCO_PATH}/${MUJOCO_ZIP}

# Mujoco for gym and mujoco_py
ENV MUJOCO_VERSION=150 \
    MUJOCO_PATH=/root/.mujoco

RUN MUJOCO_ZIP="mjpro${MUJOCO_VERSION}_linux.zip" \
    && mkdir -p ${MUJOCO_PATH} \
    && wget -P ${MUJOCO_PATH} https://www.roboti.us/download/${MUJOCO_ZIP} \
    && unzip ${MUJOCO_PATH}/${MUJOCO_ZIP} -d ${MUJOCO_PATH} \
    && rm ${MUJOCO_PATH}/${MUJOCO_ZIP}

ENV LD_LIBRARY_PATH /root/.mujoco/mjpro${MUJOCO_VERSION}/bin:${LD_LIBRARY_PATH}

COPY ./environment.yml /tmp/
COPY ./requirements.txt /tmp/
RUN conda env update -f /tmp/environment.yml \
    && rm /tmp/requirements.txt \
    && rm /tmp/environment.yml

RUN echo "source activate softlearning" >> /root/.bashrc
ENV BASH_ENV /root/.bashrc

# Trigger mujoco_py compilation using the MJKEY provided. Delete MJKEY afterwards.
RUN echo "${MJKEY}" > /root/.mujoco/mjkey.txt \
    && bash -c "source activate softlearning \
                && python -c 'import mujoco_py'" \
    && rm /root/.mujoco/mjkey.txt

COPY ./docker/entrypoint.sh /entrypoint.sh

ENTRYPOINT ["/entrypoint.sh"]