Published 4/2024
MP4 | Video: h264, 1920x1080 | Audio: AAC, 44.1 KHz
Language: English | Size: 2.06 GB | Duration: 3h 57m
Practice Large Language Models with Amazing Exercises
What you'll learn
Deep Learning
Transformers
Langchain
Large Language Models
Requirements
Python Knowledge
Basic Deep Learning Knowledge
Transformers knowledge is desirable
Description
Dive into the revolutionary world of Large Language Models (LLMs) with our comprehensive 4-hour workshop, designed to bridge the gap between theoretical knowledge and practical skills. Whether you're a budding data scientist, an AI enthusiast, or a seasoned professional looking to expand your toolkit, this course is tailored to empower you with hands-on experience in leveraging LLMs for a variety of real-world applications.What You'll Learn:Fundamentals and Advanced Techniques: Start with the basics of Large Language Models, including their architecture and capabilities, before progressing to advanced optimization methods such as Quantization and LoRA.Practical Exercises: Engage in structured exercises using Kaggle datasets in Colab, fine-tuning models for tasks like question answering and text summarization with QLoRA, and exploring cutting-edge concepts such as Retrieval Augmented Generation (RAG).Real-World Applications: Tackle engaging projects like building a semantic search engine to find movies and developing a chat interface with scholarly articles, applying your knowledge in tangible, impactful ways.Model Publication: As a bonus, learn how to share your fine-tuned models with the world through Huggingface, enhancing your visibility in the AI community.Intended Learners:This course is perfect for individuals looking to deepen their understanding of LLMs and apply these models in innovative ways. Ideal for AI professionals, data scientists, and researchers eager to expand their skills and apply LLMs to solve complex problems.
Overview
Section 1: Introduction
Lecture 1 Introduction
Lecture 2 How to use any dataset on Kaggle in Colab
Lecture 3 LLMs Training Optimization Methods - Quantization & LoRA
Lecture 4 What is RAG ?
Lecture 5 Evaluation Methods for LLMs
Section 2: Full Fine-tuning for Question Answering
Lecture 6 Introduction to the problem & the dataset
Lecture 7 Data preparation
Lecture 8 Model and tokenizer
Lecture 9 Training
Lecture 10 Evaluation and Testing
Section 3: Fine-tuning for News-Text Summarization (QLoRA)
Lecture 11 Introduction to the problem & the dataset
Lecture 12 Data preparation
Lecture 13 Tokenizer
Lecture 14 Model Preparation
Lecture 15 Training
Lecture 16 Evaluation
Lecture 17 Extra: Publish your model on Huggingface
Section 4: Find your movie - Semantic Search
Lecture 18 Introduction to the problem & dataset
Lecture 19 Data Preparation
Lecture 20 Vector Database creation
Lecture 21 Testing
Section 5: Chat with your paper - Retrieval Augmented Generation (RAG)
Lecture 22 Introduction to the problem & dataset
Lecture 23 Data Preparation & Preprocessing
Lecture 24 Vector Database creation
Lecture 25 LLM Prompt Integration & Model
Lecture 26 Testing
Deep learning Engineers,Deep Learning enthusiasts,Machine Learning Engineers,Artificial intelligence Engineers
TO MAC USERS: If RAR password doesn't work, use this archive program:
RAR Expander 0.8.5 Beta 4 and extract password protected files without error.
TO WIN USERS: If RAR password doesn't work, use this archive program:
Latest Winrar and extract password protected files without error.