警告：使用sudo CUDA_PATH安装链接器时，nvcc路径！= CUDA_PATH

Question

I was running a code in orenmel , it is cool with CPU. But when I ran -train_context2vec.py- with GPU, I recieved：

Warning: nvcc path != CUDA_PATH ... cuda.check_cuda_available() RuntimeError: CUDA environment is not correctly set up, cannot import name core

So I did this: image

Then I used this sudo CUDA_PATH=/usr/local/cuda-8.0 pip install chainer . It said:

error: command "x86_64-linux-gnu-gcc" failed with exit status 1".

If the CUDA_PATH is set somewhere else.Or I just use sudo pip install chainer.It can be installed. Others can use the gpu in Keras and theano.

Answer 1

这是通过卸载theano解决的。