A3AIARC102: AI Fine Tuning and Data Preparation for Pre-Trained Models
About this Course
You will develop the skills to gather, clean, and organize data for fine-tuning pre-trained LLMs and Generative AI models. Through a combination of lectures and hands-on labs, you will use Python to fine-tune open-source Transformer models. Gain practical experience with LLM frameworks, learn essential training techniques, and explore advanced topics such as quantization. During the hands-on labs, you will access a GPU-accelerated server for practical experience with industry-standard tools and frameworks.
Audience Profile
• Project Managers
• Architects
• Developers
• Data Acquisition Specialists
At Course Completion
• Clean and Curate Data for AI Fine-Tuning
• Establish guidelines for obtaining RAW Data
• Go from Drowning in Data to Clean Data
• Fine-Tune AI Models with PyTorch
• Understand AI architecture: Transformer model
• Describe tokenization and word embeddings
• Install and use AI frameworks like Llama-3
• Perform LoRA and QLoRA Fine-Tuning
• Explore model quantization and fine-tuning
• Deploy and Maximize AI Model Performance
Outline
Data Curation for AI
• 💬 Lecture: Curating Data for AI
• 💻 Lecture + Lab: Gathering Raw Data
• 💻 Lecture + Lab: Data Cleaning and Preparation
• 💻 Lecture + Lab: Data Labeling
• 💻 Lecture + Lab: Data Organization
• 💬 Lecture: Premade Datasets for Fine Tuning
• 💻 Lecture + Lab: Obtain and Prepare Premade Datasets
Deep Learning
• 💬 Lecture: What is Intelligence?
• 💬 Lecture: Generative AI
• 💬 Lecture: The Transformer Model
• 💬 Lecture: Feed Forward Neural Networks
• 💻 Lecture + Lab: Tokenization
• 💻 Lecture + Lab: Word Embeddings
• 💻 Lecture + Lab: Positional Encoding
Pre-trained LLM
• 💬 Lecture: A History of Neural Network Architectures
• 💬 Lecture: Introduction to the LLaMa.cpp Interface
• 💬 Lecture: Preparing A100 for Server Operations
• 💻 Lecture + Lab: Operate LLaMa3 Models with LLaMa.cpp
• 💻 Lecture + Lab: Selecting Quantization Level to Meet Performance and Perplexity Requirements
Fine Tuning
• 💬 Lecture: Fine-Tuning a Pre-Trained LLM
• 💬 Lecture: PyTorch
• 💻 Lecture + Lab: Basic Fine Tuning with PyTorch
• 💻 Lecture + Lab: LoRA Fine-Tuning LLaMa3 8B
• 💻 Lecture + Lab: QLoRA Fine-Tuning LLaMa3 8B
Operating Fine-Tuned Model
• 💬 Lecture: Running the llama.cpp Package
• 💻 Lecture + Lab: Deploy Llama API Server
• 💻 Lecture + Lab: Develop LLaMa Client Application
• 💻 Lecture + Lab: Write a Real-World AI Application using the Llama API
Prerequisites
• Python or Equivalent Experience
• Familiarity with Linux