krishnak3101/optimization-of-cnn

optimization of CNN using sparse ternary connect and dual indexing model

requirements needed to run the project code:

Anaconda

Jupyter notebook

python 3.5

Numpy

Tqdm

Few details about the code:

The code for a partular method and a particular architecture can be found at the subfolder with the name of the architecture and the file with the name of the method.every code is ran for 3 epochs .

steps to run the code:

In all the jupyter files run all the blocks until until the markdown stating 'Ignore the blocks below'. After completion of running the code the results will be found on the block above the markdown block.