Run with GPU about neural-style-transfer

[ashnakapoor07]

Can I take this up as my first issue?

[tejasa97]

@ashnakapoor07 Correct me if I'm mistaken, but I'm not sure anybody requires permission. Knock yourself out!

[titu1994]

If you have CUDA setup with Tensorflow / Theano, the scripts should automatically use the GPU.

If you do not have a local GPU, the colab link in the readme points to a GPU enabled notebook which you can use the script from, just run the cells one by one, select the content and style image and the options and there you go.

Also, @ashnakapoor07 Theres no need to request permission to resolve an issue. Go right ahead.

[f1am3d]

@tejasa97

1. Check tensorflow-gpu version that this project is using.
2. Check CUDA compatibility with tensorflow-gpu version.
3. Install needed CUDA version.
4. Check system paths variables for CUDA installation.
5. Install CUDA-compatible cuDNN version.
6. Reboot your PC.
7. Run from tensorflow.python.client import device_lib print(device_lib.list_local_devices())