Optimization and visualization of AlexNet architecture using Tensorflow 2.0 and Tensorboard.
Google Colab is used for processing while Google Drive is used for storage of datasets.
Graph visualization of loss, accuracy, and gradients for better understanding of neural network in training.
Hyperparameter optimization is also performed using Tensorboard.
Tree directory of dataset
> Train Folder
>> class 1 Folder
>>> image 1
>>> image 2
...
>> class 2 Folder
>>> image 1
>>> image 2
...
>> ...
> Test Folder
>> ...