Scene Graph Generation Note 3

本文记录一下Hiwi的相关笔记, 第三部分。

conda install pytorch=1.0
pip install cython
pip install matplotlib numpy scipy pyyaml packaging pycocotools tensorboardX tqdm pillow scikit-image
conda install opencv

envriment

• PATH
  export PATH=/home/xujing/miniconda3/envs/cl4vrd/bin:/home/xujing/miniconda3/condabin:/usr/local/cuda-10.0/bin:/home/xujing/.vscode-server/bin/58bb7b2331731bf72587010e943852e13e6fd3cf/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/snap/bin

• PYTHONPATH
  export PYTHONPATH=/home/xujing/Dokumente/ContrastiveLosses4VRD

GPSNet

https://github.com/taksau/GPS-Net/tree/master/gpsnet-detectron

export PATH=/home/xujing/miniconda3/envs/gpsnet/bin:/home/xujing/miniconda3/condabin:/usr/local/cuda-9.0/bin:/home/xujing/.vscode-server/bin/d2e414d9e4239a252d1ab117bd7067f125afd80a/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin
export PYTHONPATH=~/Dokumente/GPS-Net/gpsnet-detectron/:~/Dokumente/GPS-Net/gpsnet-detectron/lib/:~/Dokumente/GPS-Net/gpsnet-detectron/Detectron.pytorch/lib/

• 如何解决 #error – unsupported GNU version! gcc versions later than 6 are not supported!
  添加 --compiler-bindir /path/to/gcc-5 :
  eg:
  nvcc -c -o roi_crop_cuda_kernel.cu.o roi_crop_cuda_kernel.cu
  -D GOOGLE_CUDA=1 --compiler-bindir /usr/bin/gcc-5 -x cu -Xcompiler -fPIC $CUDA_ARCH

isuss

• 如何解决 /Usr/lib/x86_64-Linux-gnu/libstdc++.so.6 Version GLIBCXX_3.4.22’ Not Found

  export LD_LIBRARY_PATH=/home/xujing/miniconda3/envs/dynsg/lib/

• Q: ImportError: torch.utils.ffi is deprecated. Please use cpp extensions instead.
  A:

- from torch.utils.ffi import create_extension
+ from torch.utils.cpp_extension import BuildExtension as create_extension

• Q: ValueError: numpy.ufunc size changed, may indicate binary incompatibility. Expected 216 from C header, got 192 from PyObject
  A: conda install numpy=1.17.4

• 解决软连接参数过长

find /data/scene_understanding/Visual_Genome/VG_100K/ -name "*.jpg" | xargs -i ln -s {} VG_100K

reference

PyTorch 学习笔记（六）：PyTorch的十八个损失函数