Building-Your-Own-Neural-Network

Deep Learning Beginner Project - CIFAR-10 Image Classification

Project Overview

This project is an entry-level deep learning project for beginners. It implements the classic ResNet architecture using PyTorch to classify images in the CIFAR-10 dataset. Through this project, you will learn how to build a complete deep learning pipeline, including data loading, model definition, training, evaluation, and prediction.

Project Structure

• datasets.py: Defines the code for data loading and transformation.
• resnet_model.py: Contains the ResNet model built from scratch.
• train.py: Implements the model training process, including loss computation, backpropagation, and gradient updates.
• evaluate.py: Evaluates the model's performance on the test set.
• predict.py: Loads the trained model and makes predictions on new images.
• vis.py: Provides visualization tools to display training loss and accuracy curves.
• requirements.txt: Contains the dependencies required for the project, which can be installed using pip install -r requirements.txt.
• log/: Stores trained models.
• loss/: Stores loss and accuracy curves from the training process.

Dataset

This project uses the CIFAR-10 dataset, which contains 60,000 32x32 color images in 10 classes, with 50,000 training images and 10,000 test images.

How to Run the Project

1. Clone the Repository

   git clone https://github.com/litleWY/Building-First-Neural-Network.git
   cd Building-Your-Own-Neural-Network

2. Install Dependencies

   Install the required dependencies by running:

   pip install -r requirements.txt

3. Train the Model

   Run train.py to train the ResNet model:

   python train.py --batch_size 64 --learning_rate 0.001 --epochs 20

4. Evaluate the Model

   Run evaluate.py to evaluate the model on the test set:

   python evaluate.py

5. Predict Image Classes

   Use the trained model to predict new images:

   python predict.py

Project Features

• ResNet from Scratch: The project does not use a pre-trained model but instead builds ResNet-18 from scratch, helping you understand residual networks in deep learning.
• Training By AppleM3: Uses AppleM3 to train the model.
• Training Visualization: The vis.py script generates training loss and accuracy curves, saving them in the loss/ folder to help you understand model convergence.

Dependencies

• Python 3.6+
• PyTorch
• torchvision
• matplotlib
• numpy

Folder Description

• log/: Stores the trained model parameter files.
• loss/: Stores loss and accuracy curves generated during training.

Contribution Guidelines

Contributions are welcome! If you have any suggestions for improvements or encounter any issues, feel free to submit an issue or pull request.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgments

Thanks to the PyTorch team and the open-source community for providing excellent tools and tutorials that made this project possible.