IFT-6759-Team-13

Project 1: Advanced Projects in Machine Learning

evaluator.py run with args.

Regenerate Results:

Preprocessing:

cd prepro
python data_cleaning.py
cd ..
python -m utilities.dataloader2 --data_catalog_path='prepro/clean_df'

Training

1. Run command:

python prepro/data_cleaning_no_melt.py

This will create train_df, valid_df, and test_df

2. Run command:

python -m utilities.dataloader_simple2

It takes ~4 hours to create x,y cache. This enables very fast training and validation.

3. Run command:

python -m model.simple_keras_cnn