Tx2编译torch及libtorch

torch编译

适用于TX2 JetPack4.2，系统内置Ubuntu版本18.04，GCC版本7.4.0，CUDA版本10.0，cudNN版本7.3.1

• pytorchV1.1.0 python2.7

  wget https://nvidia.box.com/shared/static/o8teczquxgul2vjukwd4p77c6869xmri.whl -O torch-1.1.0-cp27-cp27mu-linux_aarch64.whl
  pip install torch-1.1.0-cp27-cp27mu-linux_aarch64.whl
  

• pytorchV1.1.0 python3.6

  wget https://nvidia.box.com/shared/static/j2dn48btaxosqp0zremqqm8pjelriyvs.whl -O torch-1.1.0-cp36-cp36m-linux_aarch64.whl
  pip3 install numpy torch-1.1.0-cp36-cp36m-linux_aarch64.whl
  

• pytouchV1.0.0 python2.7

  wget https://nvidia.box.com/shared/static/d5v4bngglqhdbr4g9ir4eeg6k6miwqnv.whl -O torch-1.0.0a0+bb15580-cp27-cp27mu-linux_aarch64.whl
  pip install torch-1.0.0a0+bb15580-cp27-cp27mu-linux_aarch64.whl
  

• pytorchV1.0.0 python3.6

  wget https://nvidia.box.com/shared/static/2ls48wc6h0kp1e58fjk21zast96lpt70.whl -O torch-1.0.0a0+bb15580-cp36-cp36m-linux_aarch64.whl
  pip3 install numpy torch-1.0.0a0+bb15580-cp36-cp36m-linux_aarch64.whl
  

适用于TX2 JetPack3，系统内置Ubuntu版本16.04，GCC版本5.4.0，CUDA版本9.0, python3.5
自己结合官方和一些其他资料编成功的：

# 安装依赖库
sudo apt install libopenblas-dev libatlas-dev liblapack-dev
pip3 install --upgrade pip==9.0.1
sudo pip3 install numpy scipy
sudo pip3 install pyyaml
sudo pip3 install scikit-build
pip3 install ninja --user

sudo apt-get -y install cmake
sudo apt install libffi-dev
sudo pip3 install cffi

# 添加环境变量
sudo gedit ~/.bashrc
export CUDNN_LIB_DIR=/usr/lib/aarch64-linux-gnu
export CUDNN_INCLUDE_DIR=/usr/include
source ~/.bashrc


# 下载1.0版本pytorch
git clone --recursive --depth 1 https://github.com/pytorch/pytorch.git -b v1.0.1

# 安装依赖
cd pytorch
sudo pip3 install -U setuptools
sudo pip3 install -r requirements.txt

# 开始编译
python3 setup.py build
#上面执行完再执行下面的，每个命令需要时间挺久的
NO_SYSTEM_NCCL=1 DEBUG=1 sudo python3 setup.py build develop

torch 编译（官方的，没成功可能是版本关系吧）

https://devtalk.nvidia.com/default/topic/1049071/jetson-nano/pytorch-for-jetson-nano/

$ git clone --recursive http://github.com/pytorch/pytorch  这里最好找到相应的版本，当前是V1.1.0
$ cd pytorch
$ export USE_NCCL=0
$ export USE_DISTRIBUTED=0
$ export TORCH_CUDA_ARCH_LIST="5.3;6.2;7.2

python2.7

$ sudo apt-get install python-pip cmake
$ pip install -U pip  这里升级pip到19，可能会被变成pip2了
$ sudo pip install -U setuptools
$ sudo pip install -r requirements.txt
$ pip install scikit-build --user
$ pip install ninja --user
$ python setup.py bdist_wheel

python3.6

$ sudo apt-get install python3-pip cmake
$ sudo pip3 install -U setuptools
$ sudo pip3 install -r requirements.txt
$ pip3 install scikit-build --user
$ pip3 install ninja --user
$ python3 setup.py bdist_wheel