FROM ubuntu:16.04 RUN apt update RUN apt install -y build-essential RUN apt install -y llvm-6.0 clang-6.0 RUN apt install -y git wget curl ca-certificates openssl vim zip RUN apt install -y python3 python3-dev python3-distutils RUN apt install -y libedit-dev libxml2-dev antlr4 RUN wget https://cmake.org/files/v3.22/cmake-3.22.0-linux-x86_64.sh \ && bash cmake-3.22.0-linux-x86_64.sh --skip-license --prefix=/usr/local RUN curl https://bootstrap.pypa.io/pip/3.6/get-pip.py -o get-pip.py && python3 get-pip.py && rm get-pip.py RUN pip3 install -U pip setuptools wheel RUN pip3 install numpy decorator attrs antlr4-python3-runtime RUN wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/cuda-repo-ubuntu1604_10.0.130-1_amd64.deb RUN dpkg -i cuda-repo-ubuntu1604_10.0.130-1_amd64.deb RUN apt-key adv --fetch-keys http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/7fa2af80.pub RUN apt update RUN DEBIAN_FRONTEND=noninteractive apt install -y cuda-nvcc-10-0 cuda-libraries-10-0 cuda-nvrtc-dev-10-0 cuda-cudart-dev-10-0 cuda-runtime-10-0 ENV PATH="${PATH}:/usr/local/cuda-10.0/bin" WORKDIR /root RUN git clone --recursive https://github.com/apache/incubator-tvm.git tvm WORKDIR /root/tvm/build RUN cmake -DUSE_LLVM=ON -DUSE_ANTLR=ON -DUSE_CUDA=ON .. && make -j$(nproc) WORKDIR /root/tvm/python RUN python3 setup.py install WORKDIR /root/tvm/topi/python RUN python3 setup.py install WORKDIR /root RUN pip3 install tensorflow==1.15 keras keras-applications RUN pip3 install mxnet gluoncv WORKDIR /root/ec2_compilation_container COPY tvm_ec2_compiler_utils.py compile_keras.py compile_tensorflow.py compile_gluoncv.py ./