jetson安装mmdeploy

以下是在Jetson上安装mmdeploy的步骤：
1. 创建mmdeploy环境，Python版本为3.6：

conda create -n mmdeploy python=3.6 -y
conda activate mmdeploy

2. 克隆mmdeploy代码并安装依赖：

git clone https://github.com/megvii-model/mmdeploy.git
cd mmdeploy
pip install -r requirements.txt

3. 构建mmdeploy：

python setup.py build
python setup.py install

4. 下载TensorRT并安装：

wget https://developer.nvidia.com/compute/machine-learning/tensorrt/7.2.3/local_repos/nv-tensorrt-repo-ubuntu1804-cuda11.1-trt7.2.3.4-ga-20210226_1-1_amd64.deb
sudo dpkg -i nv-tensorrt-repo-ubuntu1804-cuda11.1-trt7.2.3.4-ga-20210226_1-1_amd64.deb
sudo apt-key add /var/nv-tensorrt-repo-cuda11.1-trt7.2.3.4-ga-20210226/7fa2af80.pub
sudo apt-get update
sudo apt-get install tensorrt

5. 构建mmdeploy TensorRT插件：

cd mmdeploy
mkdir build
cd build
cmake .. -DTRT_LIB=/usr/lib/aarch64-linux-gnu/libnvinfer.so.7.2.3 -DTRT_BIN=/usr/lib/aarch64-linux-gnu/libnvinfer_plugin.so.7.2.3 -DPYTHON_INCLUDE_DIR=$(python -c "from distutils.sysconfig import get_python_inc; print(get_python_inc())") -DPYTHON_LIBRARY=$(python -c "import distutils.sysconfig as sysconfig; print(sysconfig.get_config_var('LIBDIR'))")/libpython3.6m.so
make -j$(nproc)

6. 启动tritonserver并指定预加载库：

LD_PRELOAD=/root/mmdeploy/build/libmmdeploy_tensorrt_ops.so ./tritonserver --model-repository=/root/models