On-Device Training and Transfer Learning - Lecture 16

Explore system support for efficient on-device training in this lecture from MIT's course on TinyML and Efficient Deep Learning Computing. Dive into the Tiny Training Engine and its applications for on-device training and transfer learning. Gain insights into deploying neural networks on mobile and IoT devices, as well as techniques for accelerating training processes. Learn from instructor Song Han about model compression, pruning, quantization, neural architecture search, and distillation. Discover efficient training methods, including gradient compression and on-device transfer learning. Apply knowledge to optimize models for videos, point cloud, and NLP applications. Get hands-on experience implementing deep learning applications on microcontrollers, mobile phones, and quantum machines through an open-ended design project focused on mobile AI.

Lecture 16 - On-Device Training and Transfer Learning (Part II) | MIT 6.S965