This repository showcases implementations of various deep learning neural network architectures using TensorFlow and PyTorch frameworks. The aim is to provide a comprehensive guide for learning and understanding deep learning concepts by comparing these two popular frameworks.
Before using this repository, make sure you have the following installed:
- Python 3.x
- TensorFlow
- PyTorch
- NumPy
- Matplotlib
- Other libraries listed in
requirements.txt
To get started, clone the repository and install the required dependencies:
git clone https://github.com/MDalamin5/DeepLearning-With-Tensorflow-and-Pytorch
cd repo-name
pip install -r requirements.txtThe repository contains implementations of the following neural networks:
- Feedforward Neural Networks (FNN)
- Implementations using both TensorFlow and PyTorch
- Convolutional Neural Networks (CNN)
- For image classification tasks
- Recurrent Neural Networks (RNN)
- Focused on sequential data processing
- Long Short-Term Memory (LSTM)
- Dealing with long-term dependencies in sequences
- Generative Adversarial Networks (GAN)
- Generating synthetic data from noise
- Transfer Learning
- Using pre-trained models for new tasks
- Reinforcement Learning
- Neural networks for policy learning
Each model comes with detailed Jupyter notebooks for training, evaluation, and experimentation. To run a specific model:
- Navigate to the respective directory for TensorFlow or PyTorch.
- Run the provided Python scripts or open the corresponding Jupyter notebook.
For example, to run the TensorFlow-based CNN implementation:
cd tensorflow/cnn
python train_cnn.pyFor PyTorch-based RNN:
cd pytorch/rnn
python train_rnn.pyThe implementations are inspired by the following resources:
Contributions are welcome! Please fork the repository, create a new branch, and submit a pull request. Make sure your code is well-documented and adheres to the existing style.