Tinker Workshop: Fine-Tuning Models for Fun and Profit

​(tentatively in NS Library)

Learn how to fine-tune small open-source models to replace expensive LLM API calls for production tasks like classification, extraction, and routing — cutting costs by 100x while dropping latency from seconds to milliseconds. We'll walk through the full pipeline hands-on: generating training data with a "teacher" model, LoRA fine-tuning, and evaluating your distilled model.

Laptops recommended!
/tnguyen21/lora-demo