Comprehensive Generative AI and Large Language Models - from fundamentals to production applications.
This repository documents my exploration of Generative AI, covering everything from foundational NLP concepts to modern LLM applications. Each notebook represents hands-on implementation of key GenAI techniques.
- AI Cover Letter Generator - Complete LangGraph application with Gradio UI
- AI Dad Jokes Generator - A funny experiment with OpenAi model and Gradio UI
Also deployed at Hugging Face Spaces
- Simple ChatBot - Conversational AI implementation
- LLMs with RAG and HuggingFace - Retrieval-augmented generation system
- Self-Attention & Positional Encoding - Core transformer mechanisms
- Transformer for Translation - Sequence-to-sequence models
- Transformers for Classification - Text classification tasks
- Decoder-Only Models (GPT) - Causal language modeling
- Fine-Tuning with SFT - Supervised fine-tuning
- Pre-training BERT - Masked language modeling
- Pre-training LLMs - Training from scratch
- Reward Modeling - RLHF foundations
- Data Prep for BERT - Dataset preparation
- Tokenization - BPE, WordPiece, SentencePiece
- Word2Vec - Word embeddings
- N-gram Language Models - Statistical language modeling
- RNN Translation - Sequence-to-sequence with RNNs
- HuggingFace Models - Working with pre-trained models
- LangChain Document Loader - Document processing pipelines
- NLP Data Loader - Dataset handling
Frameworks & Libraries
- Transformers (HuggingFace)
- LangChain
- LangGraph
- PyTorch
- TensorFlow
Models
- BERT, GPT, T5
- Llama, Mistral
- Custom trained models
Applications
- RAG systems
- Fine-tuning & PEFT
- Prompt engineering
- Multi-agent systems
✅ Transformer architecture and self-attention
✅ Pre-training vs fine-tuning strategies
✅ Retrieval-augmented generation (RAG)
✅ Prompt engineering and optimization
✅ RLHF and reward modeling
✅ Production deployment patterns
✅ Multi-agent LLM systems
For broader AI/ML context:
- AI-Agents - Multi-agent systems with crewAI and LangGraph
- Deep-Learning-Fundamentals - Neural network foundations
- Python-AI-Applications - Computer vision and audio processing
- ML-Fundamentals-Portfolio - Classical ML algorithms
This repository represents my deep dive into Generative AI and LLMs - the technologies reshaping AI in 2025. Each notebook combines theoretical understanding with practical implementation, suitable for both learning and reference.
Built with 22+ years of ML experience, now focused on cutting-edge GenAI applications.
All notebooks created and tested in Google Colab - ready to run.