Repository for pytorch examples and experiments.
- 1_cifar10_model_architectures: Trains the selected network architecture on the CIFAR10 dataset.
Use Visual Studio Installer -> select Modify.
Workload:
- Install Data science and analytical applications workload
Components:
- Install Python language support
- Install Python 3 64-bit (e.g. 3.6.3)
Install required packages:
- Start Visual Studio, open Python Environments
- Select Python 3.6 (64-bit)
- On the Overview combo box select Install from requirements.txt
For the gpu version, make sure to check which version of cuda and cuDNN is supported. Then follow the instructions on cudnn-install-windows