This repository contains two implementations of a simple image classification neural network using TensorFlow and PyTorch.
- Python 3.x
- TensorFlow (for TensorFlow implementation)
- PyTorch (for PyTorch implementation)
- Install the required packages:
pip install tensorflow - Run the script:
python tensorflow_impl.py
- Install the required packages:
pip install torch - Run the script:
python pytorch_impl.py
- Input shape: 256x256x3
- Convolutional layers:
- TensorFlow: 32 filters of size 3x3 and 64 filters of size 3x3, both with ReLU activation
- PyTorch: 32 filters of size 3x3 and 64 filters of size 3x3, both with ReLU activation
- Pooling layers:
- TensorFlow: Max pooling with 2x2 pool size
- PyTorch: Max pooling with 2x2 pool size
- Hidden layers:
- TensorFlow: Two fully connected layers with 128 and 64 units, both with ReLU activation
- PyTorch: Three fully connected layers with 128, 64, and 10 units, all with ReLU activation
- Output layer:
- TensorFlow: Fully connected layer with 10 units and softmax activation
- PyTorch: Fully connected layer with 10 units
- Optimizer: Adam
- Loss function: Sparse categorical cross-entropy
- Metrics: Accuracy
- Optimizer: Adam
- Loss function: Cross-entropy
- Metrics: Accuracy