https://github.com/bonifatius94/backprop
Deep Learning Reference Implementation
https://github.com/bonifatius94/backprop
deep-learning numpy reference-implementation
Last synced: 4 months ago
JSON representation
Deep Learning Reference Implementation
- Host: GitHub
- URL: https://github.com/bonifatius94/backprop
- Owner: Bonifatius94
- Created: 2023-02-19T13:22:23.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-04-07T19:06:55.000Z (about 2 years ago)
- Last Synced: 2025-01-07T20:30:46.241Z (6 months ago)
- Topics: deep-learning, numpy, reference-implementation
- Language: Python
- Homepage:
- Size: 8.79 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Backprop - Deep Learning Reference Implementation
## About
This repository outlines the math behind popular deep learning
frameworks like TensorFlow or PyTorch facilitating techniques
based on backpropagation and math calculus.
## Contents
### Deep Regression: Estimate sin(x)
- Architecture: Feed-forward neural network
- Layers: Dense, ReLU, Sigmoid
- Losses: MSE
- Optimizers: SGD, ADAM
```sh
python3 deep_regression.py
```
### Deep Classification: Categorize Hand-Written Digits
- Architecture: Feed-forward neural network
- Layers: Dense, ReLU, Sigmoid
- Losses: Cross Entropy
- Optimizers: SGD, ADAM
```sh
python3 deep_classification.py
```