This repository provides an exemplary blueprint (based on the blog post at neptune.ai for the file structure of a machine learning project, which is intended to be populated as follows but can of course be freely adapted to fit your needs:
Directory | Intended Contents |
---|---|
data | Datasets |
models | Model parameters, checkpoints etc. |
notebooks | Jupyter notebooks used e.g. for data exploration or prototyping |
reports | Experimental results, training logs, data visualizations |
src/data | Data handlers, -generators etc. |
src/models | Model implementations |
tools | Experiment scripts |
To use the blueprint, perform the following steps:
- Either download the project as a .zip file or create a fork.
- Replace this README.md with one that describes your project.
- Fill out the name, and version of your project in
setup.py
. - Add your Python dependencies to
requirements.txt
. - Create a Python environment for your project using
conda env create -n ENVNAME --file environment.yml
or alternatively use virtualenv. - Activate your environment using
conda activate ENVNAME
and start developing.