Today use Anaconda3 to install TensorFlow and Caffe, the main reference blog
Now the computer environment:
ubuntu16.04
cuda8.0
cudnn6.0
Anaconda3
1. From Scipy.misc import imread,imresize error
Hint error importerror:cannot import name Imread
But import scipy is displayed correctly.
Solution: Pip install Pillow.
2. Libcublas.so.9.0:cannot open Shared object file:no such file or directory
Cause: The new version of TensorFlow (after 1.5) does not support CUDA8.0. Either drop the Tennsorflow version or the Cuda version
Pip install tensorflow-gpu==1.4 to choose to install a specific version TensorFlow
3. Swig/python detected a memory leak of type ' int64_t * ', no destructor found "
Cause: tensorflow1.3 version itself bug
Solution: Install tensorflow1.4
4.anaconda Installation Caffe Problem
Cause: I don't know how to change makefile.config.
Solution: Follow the blog plan honestly use your computer to bring your own Python installation
5. Caffe installation completed test times error: Libcublas.so.8.0:cannot open Shared object file:no such file or directory
Cause: The cuda8.0 Library was not found and was not shared
Solution: Follow the blog scenario by entering the following command
sudo cp/usr/local/cuda-8.0/lib64/libcudart.so.8.0/usr/local/lib/libcudart.so.8.0 && sudo ldconfig
sudo cp/usr/local/cuda-8.0/lib64/libcublas.so.8.0/usr/local/lib/libcublas.so.8.0 && sudo ldconfig
sudo cp/usr/local/cuda-8.0/lib64/libcurand.so.8.0/usr/local/lib/libcurand.so.8.0 && sudo ldconfig
ubuntu16.0 Anaconda3 installation TensorFlow Keras Error Collection