Fine-Tuning LLMs: Dataset Design, LoRA, QLoRA, and Production Deployment

<p>The flagship, capstone course in this series: a deep, hands-on treatment of LLM fine-tuning that starts from weights and tokens, builds through dataset design and training data formats (SFT, DPO, RFT), covers training math and hyperparameters, and culminates in three real training paths &mdash; OpenAI hosted fine-tuning, local LoRA, and local QLoRA &mdash; including a full AWS EC2 GPU capstone project.</p>

Tools: python, openai fine-tuning, huggingface trl, peft/lora, qlora