TensorFlow is the most popular deep learning library as of now (2016).
This is designed to help those who want to familiarize themselves with TensorFlow functions. Particulary I focus on comparing TensorFlow functions with the equivalent functions in NumPy, the de facto standard numerical computation library. I believe this will help you be comfortable with TensorFlow quickly.
The basic outline will be as follows, though this is not 100% fixed.
- Constants, Sequences, and Random Values (DONE)
- Variables (Soon)
- Tensor Transformations (DONE)
- Math Part 1 (DONE)
- Math Part 2 (DONE)
- Math Part 3 (DONE)
- Strings (WIP)
- Control Flow (DONE)
- Images (WIP)
- Sparse Tensors (WIP)
To my knowledge, TensorFlow exercises are hard to find. Enjoy!