- conda create -n yummlyenv python=3.6 anaconda
- source activate yummlyenv
- conda install pip
- pip install -r requirement.txt
- First run
Yummly_Data_Training.ipynb
and generate model files. (Few Model files are too big...)
python3 yummly_cuisine_classification.py cuisine.unlabeled.json.gz
- It will ask for your choice of method:
- 1 for logistic regression,
- 2 for random forests,
- 3 for Text CNN.
- Based on the choice, it will load the selected model.
- After it will perform data loading and prediction of cuisines.
- At the end, it will generate a CSV on your local machine.