An ML pokemon classifier web application because I don't know enough about pokemon to classify them myself.
Install necessary python packages:
pip install -r requirements.txt
To train the model:
- Download and unzip the dataset into the model folder - https://techx.blob.core.windows.net/pokemon/dataset.zip
- Run the train.py file with the following parameters in the model folder:
python train.py --dataset dataset --model pokedex.model --labelbin lb.pickle
To test the model:
- Download and the weights into the model folder (if not previously trained) - https://techx.blob.core.windows.net/pokemon/pokedex.model
- Run the classify.py file with the following parameters in the model folder:
python classify.py --model pokedex.model --labelbin lb.pickle --image examples/charmander_counter.png
To run the web application:
- Download and the weights into the model folder (if not previously trained) - https://techx.blob.core.windows.net/pokemon/pokedex.model
- Run the app.py file from the root directory:
python app.py
- Go to localhost:5000 to access the application from the browser of your choice.