search

mljar / mljar-supervised

Python package for AutoML on Tabular Data with Feature Engineering, Hyper-Parameters Tuning, Explanations and Automatic Documentation
https://mljar.com
MIT License
2.99k stars 400 forks source link

Is it possible to use only essential files when saving & loading AutoML model? #661

Open HanChangHun opened 11 months ago

HanChangHun commented 11 months ago

Hello!

I want to upload the models I trained on GitHub to use them on other devices. However, MLJAR generates too many output files, and uploading them all to GitHub is a bit burdensome. Should I compress these training files using Zip for sharing? Are there any good techniques for saving models?

pplonski commented 11 months ago

Hi @HanChangHun,

Right now there is no such feature. You need to manually zip selected files, for example you can remove plots and additional files and just keep model files and framework.json files.

I would love to have such feature in the package.

HanChangHun commented 11 months ago

I appreciate your prompt response!