Generative AI Training in Delhi at Version IT

● Introduction-What We will Learn In This Course

• Getting Started With Python
• Python Basics-Syntax
• Variables In Python
• Basics Data Types In Python
• Operators In Python

• Conditional Statements (if, elif, else)
• Loops

● Lists and List Comprehension
● Tuples
● Dictionaries
● Sets

● Getting Started With Functions
● Lambda Function In Python
● Map Function In Python
● Filter Functions In Python

● Import Modules And Packages
● Standard Libraries Overview

● File Operations With Python
● Working with File Paths

● Exception Handling With try except else and finally blocks

● Classes & Objects
● Single And Multiple Inheritance
● Polymorphism
● Encapsulation
● Abstraction

● Tokenization
● Text Pre-processing
○ Stemming
○ Lemmatization
○ Stopwords
● Text Vectorization
○ Bag Of Words
○ N Grams
○ TF-IDF
● Word Embeddings
○ Word2Vec
○ CBOW
○ Skip Grams
○ GloVe
● Parts Of Speech Tagging
● Named Entity Recognition

● Welcome to the module on DL
● Introduction to DL
● Understanding Deep Learning

What is a Neuron
● Activation Functions
● Step Function
● Linear Function
● Sigmoid Function
● TanH Function
● ReLU Function
● Backpropagation vs Forward Pass
● Gradient Descent
● ANN Intuition
● ANN (Hyper Parameter Optimization)
● Step By Step Training With ANN
○ Optimizer
○ Loss Functions
○ Finding Optimal Hidden Layers And Hidden Neurons In ANN

● RNN Forward Propagation with Time
● Simple RNN Backward Propagation
● Problems With RNN
● End to End Deep Learning Projects with Simple RNN

● Why LSTM
● LSTM Architecture
● Forget Gate In LSTM
● Input Gate And Candidate Memory In LSTM
● Output Gate In LSTM
● Training Process In LSTM
● Variants Of LSTM
● GRU RNN Indepth Intuition

● LSTM and GRU End to End Deep Learning Project

● Bidirectional RNN
○ Why To Use It?
○ Advantages & disadvantages
○ Applications

● Introduction to Encoders
● Encoder Architecture
● Introduction to BERT
● BERT Configurations
● BERT Fine Tuning
● BERT Pre Training (Masked LM)
● Input Embeddings BERT
● RoBERTa
● DistilBERT
● AlBERT

● Introduction to Decoders
● Decoder Architecture
● GPT Architecture
● GPT (Masked Multi Head Attention)
● GPT Training

● Encoder and Decoder
● Indepth Intuition oF Encoder & Decoder
● Sequence to Sequence Architecture
● Problems With Encoder and Decoder

● Seq2Seq Networks
● Attention Mechanism Architecture

● What and Why To Use Transformers
● Understanding The basic Architecture of Encoder
● Self Attention Layer Working
● Multi Head Attention
● Feed Forward Neural Network With Multi Head Attention
● Positional Encoding
● Layer Normalization
● Layer Normalization Examples
● Complete Encoder Transformer Architecture
● Decoder-Plan Of Action
● Decoder-Masked Multi Head Attention
● Encoder and Decoder Multi Head Attention
● Decoder Final Linear And Softmax Layer

● What is Generative AI- AI Vs ML Vs DL Vs Generative AI
● How Open AI ChatGPt or LLama3 LLM Models are trained
● Evolution of LLM Models
● All LLM Models Analysis

● Data Preprocessing
○ Cleaning
○ embeddings
● End to end Generative AI Pipeline

● Introduction to Large Language Models & its architecture
● In depth intuition of transformer – Attention all your need paper
● How ChatGPT is trained.

● Introduction of hugging face
● Hands on Hugging Face – Transformers, HF Pipeline, Datasets, LLMs
● Data processing, tokenizing and feature extraction with hugging face
● Fine – Tuning using a pretrain models
● Hugging face API key generation
● Project: Text summarization with hugging face
● Project: Text to Image generation with LLM with hugging face
● Project: Text to speech generation with LLM with hugging face
● Huggingface Platform and its API

● Introduction to OpenAI
● What is OpenAI API and how to generate OpenAI API key?
● Local Environment Setup
● Hands on OpenAI – Chat completion API and Completion API
● Function Calling in OpenAI
● Project: Fine-tuning of GPT-3 model for text classification
● Project: Audio Transcript Translation with Whisper
● Project: Image generation with DALL-E

● Vector Databases
● Vector Index vs Vector Database
● How Vector db works
● Vector Database (Practicals)

● Complete Langchain Ecosystem
● Creating Virtual Environment
● Getting Started With Langchain And OpenAI

● Introduction & Installation and setup of langchain
● Prompt Templates in Langchain
● Chains in Langchain
● Langchain Agents and Tools
● Memory in Langchain
● Documents Loader in Langchain
● Multi-Dataframe Agents in Langchain
● How to use Hugging face Open Source LLM with Langchain
● Project: Interview Questions Creator Application
● Project: Custom Website Chatbot

● Introduction To Basic Components And Modules in Langchain
● Data Ingestion With Documents Loaders
● Text Splitting Techniques
○ Recursive Character Text Splitter
○ Character Text splitter
○ HTML Header Text Splitter
○ Recursive Json Splitter
● OpenAI Embedding
○ Ollama Embeddings
○ Huggingface Embeddings

● Introduction to open source LLMs – Llama
● How to use open source LLMs with Langchain
● Custom Website Chatbot using Open source LLMs
● Open Source LLMs – Falcon

● Introduction & Importance of RAG
● RAG Practical demo
● RAG Vs Fine-tuning
● Build a Q&A App with RAG using Gemini Pro and Langchain
● Retrieval Augmented Generation (RAG)

● What is fine tuning?
● Parameter Efficient Fine – Tuning
○ LoRA
○ OLoRA
○ Meta Llama 2 on Custom Data

● Introduction to LlamaIndex & end to tend Demo
● Project: Financial Stock Analysis using LlamaIndex

● How to Deploy Generative AI Application
○ Flask
○ AWS