Installation on Windows


We tested on Windows10 64bit, python3.10.

The following software are required for installation:

  • Required software.

    • Python>=3.8: PIP

  • Recommended.

    • CUDA Toolkit and cuDNN (if you have CUDA GPUs).

Setup environment


If you don’t install python, you can refer to the link below.


If your network is using proxy and setup fails, configure proxy server with environment variable and try install again.

> set http_proxy=http://(enter the address of the http proxy server here)
> set https_proxy=http://(enter the address of the https proxy server here)

CUDA and cuDNN library

If you are using a NVIDIA GPU, execution speed will be drastically improved by installing the following software.

For the versions of CUDA/cuDNN we support, see the table.

CUDA Toolkit


To install cuDNN, copy bin, include and lib to C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v{CUDA_VERSION}


See NNabla package installation using PIP.


Q. Scikit-image installation takes a long time.

Depending on the environment, it will take a long time. Please wait.