Report For Question 1

Data Exploration

Data Visualization

Data Analysis

Hyperparameter Tuning

The yellow and the green hazy lines are the ones where the learning rate is higher and that's why the Accuracy and Loss both for the training and validation are not settling since the gradient change is too much that it passes the actual minima.
Best hyperparameters are : Learning rate of 0.0005 or 0.001 where the model converges the fastest