This is the modified repository for Sara's Capstone project in an excerise to add CUDA functionality.
notes on cuda: There is very little work needed to be done to accomdiate CUDA. It is mostly hardware and driver based.
See my guide here on medium on how to get CUDA working on an Ubuntu 20.04 machine:
Actually Getting CUDA to Accelerate your Data Science for Ubuntu 20.04
Current Work: Get Keras CNN to work on CUDA. See model-1-cnn-cura.py
Next Steps: Optimize CNN structure for better optimization during validation. Current model shows a validation success rate around 25 %.\
Steps after that:
combine into a class and move away from scripts and into functions that can be accessed by main.py,
make into python module,
add pytorch implementation