Welcome to the Large Language Models section of the AI Engineering Academy. This module provides a comprehensive understanding of LLMs and their practical applications in AI engineering.
Category | Topic | Resource |
---|---|---|
Introduction | Overview | Introduction to LLMs |
Theory Behind Fine-tuning | Pre-Training | Pre-Training |
Supervised Fine-Tuning (SFT) | SFT Theory | |
Proximal Policy Optimization (PPO) | PPO Theory | |
Direct Preference Optimization (DPO) | DPO Theory | |
Observation-Regularized Policy Optimization (ORPO) | ORPO Theory | |
Gated Regularized Policy Optimization (GRPO) | GRPO Theory | |
Hands-On SFT | Overview | SFT Implementation Guide |
Implementation | SFT Notebook | |
Hands-On GRPO | Guide | Hacker Guide to GRPO |
Implementation | Qwen 0.5B GRPO | |
Gemma | Overview | Gemma Guide |
Implementation | Gemma Fine-tuning | |
Llama2 | Overview | Llama2 Guide |
Implementation | Llama2 Fine-tuning | |
Advanced | QLora Fine-tuning | |
Llama3 | Implementation | Llama3 Fine-tuning |
Mistral-7B | Overview | Mistral Guide |
Implementation | Mistral Fine-tuning | |
Evaluation | Evaluation Harness | |
DPO | DPO Fine-tuning | |
SFT | SFT Trainer | |
Inference | ChatML Inference | |
Mixtral | Implementation | Mixtral Fine-tuning |
Visual Language Models | Florence2 | Florence2 Fine-tuning |
PaliGemma | PaliGemma Fine-tuning | |
Architecture | Parameter Analysis | Parameter Count |
Level | Steps | Resources |
---|---|---|
Beginner | 1. Introduction to LLMs | Introduction |
2. Understanding core theory | Pre-Training, SFT Theory | |
3. First implementation | SFT Guide | |
4. Practical application | Llama2 Fine-tuning | |
Intermediate | 1. Advanced techniques | DPO Theory, PPO Theory |
2. Model implementation | Mistral Fine-tuning | |
3. Architecture concepts | Parameter Count | |
Advanced | 1. Cutting-edge methods | ORPO Theory, GRPO Theory |
2. Advanced implementation | GRPO Implementation | |
3. Multimodal models | Florence2, PaliGemma |
We welcome contributions to expand this repository. Please follow the standard pull request process and ensure your contributions align with the overall structure.
This project is licensed under the MIT License - see the LICENSE file for details.
AI Engineering Academy - Advancing the frontier of language model understanding and implementation