Certificate Authentication

Day 1: Fundamentals and Image Processing
- Morning Session:
● Welcome and Introduction
○ Course objectives
○ Overview of deep learning and PyTorch
● PyTorch Fundamentals
○ Introduction to Tensors
○ Basic operations with Tensors
○ Automatic differentiation with autograd
○ Hands-on Exercise: Tensor manipulations and basic operations
● Building Neural Networks with PyTorch
○ Introduction to torch.nn module
○ Creating custom neural networks using nn.Module
○ Training loops, loss functions, and optimizers
○ Hands-on Exercise: Build and train a simple feedforward neural network on a
small dataset (e.g., MNIST)
- Afternoon Session:
● Convolutional Neural Networks (CNNs)
○ Basics of CNNs
○ Layers in CNNs (Conv layers, pooling layers, etc.)
○ Implementation of a CNN in PyTorch
○ Hands-on Exercise: Build and train a CNN on an image classification task (e.g.,
CIFAR-10)
● Transfer Learning and Fine-Tuning
○ Introduction to transfer learning
○ Fine-tuning pre-trained models in PyTorch
○ Hands-on Exercise: Use a pre-trained model from torchvision.models for an
image classification task
 

Day 2: NLP with RNNs, LSTMs, GRUs, Transformers & Generative Models
- Morning Session:
● Recurrent Neural Networks (RNNs), LSTMs, and GRUs
○ Introduction to sequential data
○ Understanding RNNs, LSTMs, and GRUs
○ Implementation of simple RNN /LSTM/GRU models using PyTorch
○ Hands on Exercise : Build & train an RNN/LSTM/GRU model for text generation
or sequence prediction
● Transformers & Attention Mechanism
○ Introduction to Attention Mechanism & Transformers architecture (BERT, GPT,
etc.)
○ Implementation examples of transformers using HuggingFace’s transformers
library
○ Hands-on Exercise: Build and train a simple transformer-based model on text
classification or named entity recognition task.
- Afternoon Session:
● Generative Adversarial Network (GAN)
○ Introduction to GAN architecture (generator/discriminator)
○ Different variants of GAN (CycleGAN, WGAN, etc.)
○ Implementation of a basic GAN in PyTorch
○ Hands-on Exercise: Build and train a simple GAN for image generation (using
MNIST/CIFAR).
● GPT Architectures and Generative AI
○ Overview of GPT architecture (GPT, GPT-2, GPT-3)
○ Applications of generative AI models like ChatGPT, DALL-E
○ Hands-on Demo using OpenAI's GPT API for text generation tasks