iconismo / data-fusion-1-baseline

Slightly extended baseline solution

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

data-fusion-1-baseline by

Dmitry Dryomov @ Lamoda DS team

How to run:

  • Copy train dataset in data/data_fusion_train.parquet
  • Open full-baseline.ipynb and press "Run all"
  • Resulted submission should be at t1_sub/submission.zip
  • Baseline score is 0.8425

Most important:

  1. Full baseline including zipping of solution
  2. Validation showing score close to public LB
  3. Handling code shared between train and inference
  4. Local submit testing

About

Slightly extended baseline solution

License:MIT License


Languages

Language:Jupyter Notebook 90.5%Language:Python 9.5%