Image classification using deeper networks like ResNet, VGG, GoogLeNet etc.,
- Run
pip install -r requirements.txt
from the project directory to install the requirements. - Data will be stored in
wandb
, so it should be installed and account has to be created. Also get the api key from wandb account online and addWANDB_API_KEY
to the environment variables.
-
To
train
the model, runpython main.py train -r <data_path> -d <dataset_name> -m <model_name> -s <model_save_path>
Models Implemented: resnet18, resnet34, resnet50, resnet101, resnet152, vgg11, vgg13, vgg16, vgg19, GoogLeNet
Datasets Included: mnist, cifar10, cifar100
Note: Download MNIST dataset manually and place it in the data_path as it's website is closed down. -
To
test
the model on the test data, runpython main.py test -r <data_path> -d <dataset_name> -m <model_name> -s <model_save_path>
Models Implemented: resnet18, resnet34, resnet50, resnet101, resnet152, vgg11, vgg13, vgg16, vgg19, GoogLeNet
Datasets Included: mnist, cifar10, cifar100