Skip to content
/ aaamlp-starter Public template

Starter kit for projects from Approaching Almost Any Machine Learning problem book by Abhishek Thakur

License

Notifications You must be signed in to change notification settings

bakery/aaamlp-starter

Folders and files

NameName
Last commit message
Last commit date

Latest commit

1a4c123 · Feb 15, 2023

History

4 Commits
Feb 14, 2023
Feb 14, 2023
Feb 14, 2023
Feb 14, 2023
Feb 14, 2023
Feb 14, 2023
Feb 14, 2023
Feb 14, 2023
Feb 15, 2023

Repository files navigation

AAAMLP starter

Starter kit for projects from Approaching Almost Any Machine Learning problem book by Abhishek Thakur


Note 👷

This starter is very much work in progress. I am updating key modules as I am making my way through the book.

Prereqs

  • install miniconda
  • conda create -n NAME_FOR_YOUR_CONDA_ENV python=3.7.6 (if you are on M1/M2, you might need to run this before conda config --env --set subdir osx-64)
  • conda activate NAME_FOR_YOUR_CONDA_ENV
  • conda env create -f environment.yml
  • OR conda env create -f environment_osx.yml on Mac
  • conda activate ml
  • download mnist_train.csv from Abhishek's Kaggle, and save it to input directory

Getting started

This template gets you started with a basic skeleton for your ML app.

Creating folds

  • cd src && python create_folds.py --folds 10 (default is 5 folds)

Adding/editing available models

model_dispatcher.py contains a dictionary with models available in your application

Training your model

To train your model, use

python train.py --fold 0 --model desision_tree_entropy

About

Starter kit for projects from Approaching Almost Any Machine Learning problem book by Abhishek Thakur

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published