MNIST Neural Network in Rust
3-layer neural network trained on MNIST handwritten digit classification, built from scratch in Rust.
Model Details
- Architecture: Input (784) → Hidden (128, ReLU) → Output (10, Softmax)
- Training: 10 epochs, learning rate 0.1, batch size 32, cross-entropy loss
- Accuracy: ~97.5% on test set
Training Data
MNIST dataset — 60,000 training images, 10,000 test images of 28×28 grayscale handwritten digits.
Results
Epoch 1: ~91% accuracy
Epoch 5: ~98% accuracy
Epoch 10: ~99% train / 97.5% test
Usage
./download_mnist.sh
cargo run --release
