Introduction to Deep Learning using Pytorch with Spyder/Jupyter Notebooks for Facebook-Udacity Secure and Private AI Scholarship Challenge Program #1
Use PyTorch for building deep learning models.
- Intro to PyTorch with Tensors (main data structure of PyTorch)
- Perform simple operations to Tensors - multi-dimensional arrays
- Use PyTorch to build network and run data forward through it
- Define a loss and optimization method to train neural network on dataset of handwritten digits
- Test network is able to generalize through validation
- Use pre-trained networks to improve performance of classifier (transfer learning)