It provides modular components for models, training pipelines, evaluation, and experiment management, making it easier to develop reproducible machine learning experiments and academic prototypes.
- PyTorch-based training and evaluation pipelines
- Modular architecture for models and datasets
- Reproducible experiment structure
- Built-in unit testing with pytest
- Continuous integration with GitHub Actions
- Academic paper scaffold for research projects
- Example MNIST benchmark training pipeline
- Command-line interface for quick experimentation
Clone the repository and install in editable mode:
git clone https://github.com/Festus0/dnnkit.git
cd dnnkit
pip install -e .
pip install -r requirements-dev.txt
Command Line Interface
Train a model using the CLI:
dnnkit train --epochs 3 --dataset mnist --batch-size 64 --lr 0.001
Example Training Curve
Project Structure
dnnkit/
├── dnnkit/
│ ├── models/
│ ├── datasets/
│ ├── training/
│ └── evaluation/
│
├── tests/
├── docs/
├── scripts/
└── examples/
Development
Run tests locally:
pytest
Continuous integration runs automatically via GitHub Actions.
Roadmap
Planned improvements include:
Additional benchmark datasets
Hyperparameter search utilities
Experiment tracking integration
Model zoo
Distributed training support
License
This project is licensed under the MIT License.
Citation
If you use this repository in academic work, please cite:
Slade, F. (2026). DNNKit: A lightweight framework for reproducible deep learning experiments.