Here are some of my notes from the fifth lesson of the Fast AI course.
This lesson took a deep-dive into the inner workings of back propogation, used in previous lessons. I’m still trying to get my mind wrapped around the power of this step-wise method to exhibit such emergent behavior. The FastAI and PyTorch libaries were used to good effect to demonstrate the functioning internals.
With that knowledge, the instructor took another look at various neural nets, demonstrating the use of backpropogation within.
- This is complicated stuff.
- I need more practice…