PyTorch Tutorials

This repository contains a collection of Jupyter notebooks demonstrating core concepts and practical implementations in PyTorch. The tutorials cover everything from basic tensor operations to advanced neural network architectures and transfer learning.

Contents

• pytorch_tut_1_create_tensors.ipynb
  Introduction to creating and manipulating tensors in PyTorch.

• pytorch_tut_2_autograd.ipynb
  Understanding PyTorch's autograd system for automatic differentiation.

• Pytorch_Activation_Fn.ipynb
  Usage of activation functions (ReLU, Tanh, LeakyReLU, Sigmoid, Dropout) in neural networks.

• Pytorch_Feed_Forward_NN.ipynb
  Building and training a feed-forward neural network on the MNIST dataset.

• Pytorch_CNN.ipynb
  Implementing a Convolutional Neural Network (CNN) for image classification on CIFAR-10.

• Pytorch_RNN.ipynb
  Character-level RNN from scratch for name classification by language.

• Pytorch_RNN_GRU_LSTM.ipynb
  Building and comparing LSTM and GRU models for sequence modeling on MNIST.

• Pytorch_Softmax_&_Cross_Entropy.ipynb
  Explanation and implementation of softmax and cross-entropy loss for classification tasks.

• Pytorch_Save_Model.ipynb
  Saving and loading PyTorch models: full model, weights only, and checkpoints.

• Pytorch_Transfer_Learning.ipynb
  Transfer learning and fine-tuning using pre-trained ResNet on CIFAR-10.

Getting Started

1. Clone the repository:

   git clone https://github.com/Devguru-codes/Pytorch_Tutorials.git
   cd Pytorch_Tutorials

2. Install dependencies:

   • Python 3.x
   • PyTorch
   • torchvision
   • numpy
   • matplotlib
   • Jupyter Notebook or JupyterLab

   Install with pip:

   pip install torch torchvision numpy matplotlib notebook

3. Run the notebooks:

   jupyter notebook

   Open any notebook to explore the tutorials.

Structure

Each notebook is self-contained and focuses on a specific PyTorch concept or model. The code is well-commented for educational purposes.

License

This project is for educational purposes.