Hey Devs! ๐
I recently built a small open-source Python package called MLFCrafter that helps you build and run ML pipelines faster โ no more repetitive boilerplate!
Instead of writing tons of lines to load, clean, scale, train, and evaluate a model, you just chain crafters together.
๐ What is MLFCrafter?
MLFCrafter is a modular machine learning pipeline tool.
Each step in the pipeline is called a crafter, and you can compose them like building blocks.
Hereโs how it looks in action:
from mlfcrafter import MLFChain, DataIngestCrafter, CleanerCrafter, ScalerCrafter, ModelCrafter, ScorerCrafter, DeployCrafter
chain = MLFChain(
DataIngestCrafter(data_path="data/iris.csv"),
CleanerCrafter(strategy="auto"),
ScalerCrafter(scaler_type="standard"),
ModelCrafter(model_name="random_forest"),
ScorerCrafter(),
DeployCrafter()
)
results = chain.run(target_column="species")
print(f"Test Score: {results['test_score']:.4f}")
Thatโs it โ your full ML pipeline is up and running.
โ๏ธ Features
๐ Load your own dataset easily
๐งผ Clean and scale with customizable settings
๐ค Train a model (logistic regression, random forest, etc.)
๐ Get evaluation metrics
๐ Deploy your pipeline
๐งช Automate versioning + publishing with GitHub Workflows
๐ Easy to extend with your own custom crafters
๐ฆ Installation
pip install mlfcrafter
๐ Links
GitHub: https://github.com/brkcvlk/MLFCrafter
PyPI: https://pypi.org/project/mlfcrafter
๐ Letโs Connect
This is still early stage and Iโd love feedback, ideas, or contributions.
If you try it out and hit any issues โ Iโm here to help!
Thanks for reading! ๐
Feel free to drop a comment if you find this useful โค๏ธ
Top comments (0)