MNIST simple CNN classifier implementation in pytorch for DTU Machine Learning Operations (MLOps) course.
Start by cloning or downloading this repository
git clone https://github.com/ndpooja/MLOps_mnist.git
for project preparation
cd MLOps_mnist
make data # runs the make_dataset.py file, try it!
make clean # clean __pycache__ files
make requirements # install everything in the requirements.txt file
to train the model
make train
to predict the model
make predict
to visualize intermediate layer
python mnist_classifier/visualizations/visualize.py
The directory structure of the project looks like this:
โโโ Makefile <- Makefile with convenience commands like `make data` or `make train`
โโโ README.md <- The top-level README for developers using this project.
โโโ data
โ โโโ processed <- The final, canonical data sets for modeling.
โ โโโ raw <- The original, immutable data dump.
โ
โโโ docs <- Documentation folder
โ โ
โ โโโ index.md <- Homepage for your documentation
โ โ
โ โโโ mkdocs.yml <- Configuration file for mkdocs
โ โ
โ โโโ source/ <- Source directory for documentation files
โ
โโโ models <- Trained and serialized models, model predictions, or model summaries
โ
โโโ notebooks <- Jupyter notebooks.
โ
โโโ pyproject.toml <- Project configuration file
โ
โโโ reports <- Generated analysis as HTML, PDF, LaTeX, etc.
โ โโโ figures <- Generated graphics and figures to be used in reporting
โ
โโโ requirements.txt <- The requirements file for reproducing the analysis environment
|
โโโ requirements_dev.txt <- The requirements file for reproducing the analysis environment
โ
โโโ tests <- Test files
โ
โโโ mnist_classifier <- Source code for use in this project.
โ โ
โ โโโ __init__.py <- Makes folder a Python module
โ โ
โ โโโ data <- Scripts to download or generate data
โ โ โโโ __init__.py
โ โ โโโ make_dataset.py
โ โ
โ โโโ models <- model implementations, training script and prediction script
โ โ โโโ __init__.py
โ โ โโโ model.py
โ โ
โ โโโ visualization <- Scripts to create exploratory and results oriented visualizations
โ โ โโโ __init__.py
โ โ โโโ visualize.py
โ โโโ train_model.py <- script for training the model
โ โโโ predict_model.py <- script for predicting from a model
โ
โโโ LICENSE <- Open-source license if one is chosen
Created using mlops_template, a cookiecutter template for getting started with Machine Learning Operations (MLOps).