This project is based on kaggle's competition to correctly classify image as dog or cat.
Size of training dataset is 20000 and size of validation dataset is 5000.
Pre-trained model Resnet50 with 50 layers is used for training dataset Resnet50 gives us amazing accuracy of 0.97 ( reducing computational and time cost) with only just Epoch.
Model is saved in cloud ( in drive) for later use.
At the end submission.csv is created containing all the predictions for unseen data
if you like my approach you can thank my efforts by paying usdt
Usdt : 0x3920cdea87ba6b272f0cb41b2ddf84cf6a2c8a18 ( Bep 20 network)