Data Science Connect 11

This is a repository for Data Science Connect 11.
The notebook showcases the strengths of transfer learning and data augmentation.

Setup (Windows 64-bit)

Install conda at https://docs.conda.io/projects/conda/en/latest/user-guide/install/
Clone or download this repository.
Navigate to this the project directory in command prompt.

conda create --name <any_environment_name> --file conda_requirements.txt 
conda activate <environment_name>
jupyter notebook

Trying with your own data

1. Delete everything in data folder.
2. Create a folder for each class and place the images in the corresponding classes e.g. cat images in cat folder.
3. Place your class folders in data folder.
4. Run the notebook.
5. Enjoy!

Colab notebook (More convenient)

https://colab.research.google.com/drive/1CP668xo1Zsah3EMO3uXZeiAzyb3vSMTq