加入 Gitee
与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :)
免费加入
文件
克隆/下载
Dockerfile 2.46 KB
一键复制 编辑 原始数据 按行查看 历史
FROM nvidia/cuda:9.0-cudnn7-devel-ubuntu16.04
RUN rm -f /usr/lib/x86_64-linux-gnu/libnccl_static.a \
/usr/lib/x86_64-linux-gnu/libcudnn_static_v7.a
# Install package dependencies
RUN apt-get update && apt-get install -y --no-install-recommends \
autoconf \
automake \
libtool \
pkg-config \
ca-certificates \
wget \
git \
python \
python3 \
python-dev \
python3-dev \
libprotobuf-dev \
protobuf-compiler \
cmake \
swig \
&& rm -rf /var/lib/apt/lists/*
# Install pip
RUN cd /usr/local/src && \
wget https://bootstrap.pypa.io/get-pip.py && \
python2 get-pip.py && \
pip2 install --upgrade pip && \
python3 get-pip.py && \
pip3 install --upgrade pip && \
rm -f get-pip.py
# Build and install onnx
RUN cd /usr/local/src && \
git clone --recurse-submodules https://github.com/onnx/onnx.git && \
cd onnx && \
git checkout dee6d89 && \
pip2 install pybind11 && \
pip2 install protobuf && \
pip2 install numpy && \
pip3 install numpy && \
python setup.py build && \
python setup.py install && \
cd ../ && \
rm -rf onnx/
# Install TensorRT
WORKDIR /usr/local/src
ENV TENSORRT_VERSION 3.0
COPY TensorRT-${TENSORRT_VERSION}.*.tar.gz .
RUN tar -xvf TensorRT-${TENSORRT_VERSION}.*.tar.gz && \
cd TensorRT-${TENSORRT_VERSION}.* && \
cp lib/* /usr/lib/x86_64-linux-gnu/ && \
rm /usr/lib/x86_64-linux-gnu/libnv*.a && \
cp include/* /usr/include/x86_64-linux-gnu/ && \
cp bin/* /usr/bin/ && \
mkdir /usr/share/doc/tensorrt && \
cp -r doc/* /usr/share/doc/tensorrt/ && \
mkdir /usr/src/tensorrt && \
cp -r samples /usr/src/tensorrt/ && \
pip2 install python/tensorrt-${TENSORRT_VERSION}.*-cp27-cp27mu-linux_x86_64.whl && \
pip3 install python/tensorrt-${TENSORRT_VERSION}.*-cp35-cp35m-linux_x86_64.whl && \
pip2 install uff/uff-*-py2.py3-none-any.whl && \
pip3 install uff/uff-*-py2.py3-none-any.whl && \
cd ../ && \
rm -rf TensorRT-${TENSORRT_VERSION}.*
# Build the library
ENV ONNX2TRT_VERSION 0.1.0
WORKDIR /opt/onnx2trt
COPY . .
RUN rm -rf build/ && \
mkdir build && \
cd build && \
cmake .. && \
make -j$(nproc) && \
make install && \
ldconfig && \
cd .. && \
python setup.py build && \
python setup.py install && \
rm -rf ./build/
WORKDIR /workspace
RUN cp /opt/onnx2trt/onnx_backend_test.py .
RUN ["/bin/bash"]
Loading...
马建仓 AI 助手
尝试更多
代码解读
代码找茬
代码优化