AI
Mastering Generative AI 2025
This 4-month in-depth course will take you through the fundamentals and advanced concepts of Generative AI, covering the latest advancements in AI technologies, deep learning models, and practical applications. Students will gain hands-on experience in building and deploying AI models, and learn how to harness the power of AI in real-world scenarios, including content generation, NLP, and more.
What you will learn
- Generative AI Basics
- Understanding Generative Models
- Transformer Models and Attention Mechanism
- Deep Learning Architectures (CNNs, RNNs, GANs)
- Natural Language Processing with GPT and BERT
- Generative Text Models
- Image Generation with DALL-E
- Audio and Speech Generation with OpenAI
- Fine-tuning Pre-trained AI Models
- Building Custom AI Applications
- Advanced Applications of Generative AI (Text-to-Image, Text-to-Speech, etc.)
- AI Ethics and Responsible AI Design
Corporate training outcomes
- Completion certificate
- Practical assignments and project work
- Mentor support and progress tracking
- Custom batch options for teams
- Yes
Requirements
- Laptop with internet access
- Basic understanding of machine learning and neural networks
- Familiarity with Python programming language
- Curiosity to explore AI-driven content creation
- Willingness to experiment with real-world AI models and data
Curriculum
Introduction to Generative AI
- What is generative AI ?
- Why are generative models required?
- Understanding generative models and their significance
- Generative AI v/s Discriminative Models
- Recent advancements and research in generative AI
- Generative AI end-to-end project lifecycle
- Key applications of generative models
Text Preprocessing and Word Embedding
- Segmentation and Tokenization
- Change Case, Spell Correction
- Stop Words Removal, Punctuations Removal, Remove White spaces
- Stemming and Lemmatization
- Parts of Speech Tagging
- Text Normalization
- Rephrase Text
- One hot encoding, Index-based encoding
- Bag of words
- TF-IDF
- Word2Vec
- FastText
- N-Grams
- Elmo
- Bert-based encoding
Introduction to Large Language Models
- In-depth intuition of Transformer-Attention all your need Paper
- Guide to complete transformer tree
- When to use which transformer architecture
- Application and use cases of LLMs
- Transfer learning in NLP
- How to use pre-trained transformer-based models
- How to perform finetuning of pre trained transformer based models
- Mask language modeling
- BERT- Google
- GPT- OpenAI
- T5- Google
- Megatron- NVIDIA
- Evaluations Matrixs of LLMs models
- GPT-3 and 3.5 Turbo use cases
- Learn how Chatgpt trained
- Introduction to Chatgpt- 4
Introduction to Huggingface And its Applications
- Why the need for a hugging face?
- Introduction of Hugging Face Transformers
- Hugging face API key generation
- Hugging Face Transfer learning models based on the state-of-the-art transformer architecture
- Fine-tuning using a pre-train models
- Ready-to-use datasets and evaluation metrics for NLP.
- Data Processing, Tokenizing and Feature Extraction with Hugging Face
- Standardizing the Pipelining
- Training and callbacks
- Language Translation with Hugging Face Transformer
- Generative AI with LLMs and LLM Powered Applications
- Project: Text summarization with hugging face
- Project: Language Translation with Hugging Face Transformer
- Project: Text to Image Generation with LLM with hugging face
- Project: Text to speech generation with LLM with hugging face
Guide to Open AI and its Ready to Use Models with Application
- Introduction to OpenAI
- What is OpenAI API and how to generate OpenAI API key?
- Installation of OpenAI package
- Experiment in the OpenAI playground
- How to setup your local development environment
- Different templates for prompting
- OpenAI Models GPT-3.5 Turbo DALL-E 2, Whisper, Clip, Davinci and GPT-4 with practical implementation
- OpenAI Embeddings and Moderation with Practical Implementation
- Implementation of Chat completion API, Functional calling and Completion API
- How to manage the Tokens
- Different Tactics for getting an Optimize result
- Image Generation with OpenAI LLM model
- Speech to text with OpenAI
- Use of Moderation for content complies with OpenAI
- Understand rate limits, error codes in OpenAPI
- OpenAI plugins connect ChatGPT to third-party applications.
- How to do fine-tuning with custom data
Prompt Engineering Mastering with OpenAI
- Introduction to Prompt Engineering
- Different templates for prompting
- Prompt Engineering: What & Why?
- Prompt Engineering & ChatGPT Custom Instructions
- The Core Elements Of A Good Prompt
- Which Context Should You Add?
- Zero- One- & Few-Shot Prompting
- Using Output Templates
- Providing Cues & Hints To ChatGPT
- Separating Instructions From Content
- Ask-Before-Answer Prompting
- Perspective Prompting
- Contextual Prompting
- Emotional Prompting
- Laddering Prompting
- Using ChatGPT For Prompting
- Find Out Which Information Is Missing
- Self-evaluative Prompting
- ChatGPT-powered Problem Splitting
- Reversing Roles
- More Prompts & Finding Prompt Inspirations
- Super Prompts Like CAN & DAN
Vector database with Python for LLM Use Cases
- Introduction to vector database
- Vector database foundation
- Vector database use cases
- Text embedding
- Vector similarity search
- SQLite database
- Storing and retrieving vector data in SQLite
- Chromadb local vector database part1 setup and data insertion
- Query vector data
- Fetch data by vector id
- Database operation: create, update, retrieve, deletion, insert and update
- Application in semantic search
- Building AI chat agent with langchain and openai
- Weviate Vector Database
- Pinecone Vector Database
Hands-on with LangChain
- Introduction to langchain
- How Does LangChain Work
- Installation and setup of langchain in local env
- Hello world of LangChain application - Chaining a simple prompt
- Components of langchain like Schema, Model I/O, Prompts, Indexes, Memory, Chains, Agents, Callbacks
- Understanding prompts, language model and Output parser
- Concept of async API, fake LLM human input, LLM Caching
- Implementation of Chat models with human input chat model, chain, prompt and streaming
- Implementation of output parser with JSON parser, XML parser, and list parser
- Implement retrieval with document loader document transformer text embedding and vector store
- Implement memory with chat messages, with the conversational knowledge base, and with vector store
- Text summarization with langchain
- Question Answer with langchain
- Chatbot with LangChain
- LangChain streaming
- Embeddings and Vector Data Stores in LangChain
Hands on with LangChain
- Hands on with LangChain
- Understanding PromptTemplate + LLM + OutputParser
- LangChain expression language
- Bind runtime args
- Configurable alternatives
- Add fallback
- Run arbitrary functions
- Use RunnableParallel/RunnableMap
- Route between multiple Runnables
- Document Loaders
- CSV, PDF and JSON file analysis using LangChain
- Prompt Templating and prompt management
- Retrieval-augmented generation chain
- Multiple chains
- Querying a Sql DB
- How to add in moderation around your LLM application.
- Hugging Face Models with LangChain
- Falcon 7B fine tune on custom dataset
- Mistral 7B - Finetune and Inference for Custom Usecase
- LangChain with Google PaLM2 Model
- LangChain with Facebook Llama2 Model
- LangChain webapp with Streamlit and flask
- Project: MCQ Quiz Creator Application
- Project: Youtube video summarizer and youtube script writing
- Project: Custom Chatbot for any website
Practical Guide to LlamaIndex with LLMs
- Introduction to LlamaIndex
- Difference between langchain and LlamaIndex
- Difference between Llama and LlamaIndex
- Setup of LlamaIndex in our local env
- How to use LLMs with LlamaIndex
- Exploring Llamahub
- How to connect with external Data
- What is in Context Learning & Fine Tuning
- Why indexing required in LLM apps
- Persist indexes
- How to index our data
- Creating documents objects
- Different Documents Loader
- How to verify sources of the response
- How to connect with different documents like csv,txt,pdf etc
- Document Management
- Recursive file processing from directory sub directory
- Building apps with LlamaIndex
- Customization LLM Models in Application
- Integration with endpoint flask and streamlit
- Enable Streaming response
- Chat engine: Condense mode
- Chat engine: React mode
- Customizing Prompt
- How to use vector databases like ChromaDB and Weviate with LlamaIndex
- Token Prediction & Cost Analysis
- Integrations with OpenAI, Hugging Face
- Project: Financial Stock Analysis using LlamaIndex
- Project: Chat with Books and PDF Files with Llama 2
End to End Projects
- Industry Projects