Back to NewsUnsloth founder's 2h42m fine-tuning masterclassFederico UlfoJuly 14, 2026NewsUnsloth founder's 2h42m fine-tuning masterclass — Ex-NVIDIA engineer explains RL, kernels, reasoning, quantization, agents - better than $5,000 bootcamps. Sources: @h100envy RelatedTinyLoRA: Fine-Tuning 8B Models by Tweaking Just 13 ParametersGoodfire kills a model's German by tuning one scalar on a single weight subcomponentAbout the AuthorsFederico UlfoCo-FounderPrevSatya Nadella: The Reverse Reference ParadoxNext46 thoughts on the near future
Federico Ulfo
