Explore on-device learning techniques for IoT devices with limited memory in this tinyML Forum talk by MIT EECS Assistant Professor Song Han. Discover an algorithm-system co-design framework enabling neural network fine-tuning with only 256KB of memory. Learn about quantization-aware scaling to stabilize quantized training and sparse update techniques to reduce memory footprint. Understand the implementation of Tiny Training Engine, a lightweight system that prunes backward computation graphs and offloads runtime auto-differentiation. Gain insights into practical solutions for on-device transfer learning in visual recognition tasks on microcontrollers, using significantly less memory than existing frameworks while matching cloud training accuracy. Delve into topics such as edge learning challenges, memory bottlenecks, parameter-efficient transfer learning, and deep gradient compression. See how this approach enables continuous adaptation and lifelong learning for tiny IoT devices, demonstrating their capability to perform both inference and on-device learning.
On-Device Learning Under 256KB Memory - Challenges and Solutions for IoT Devices
Overview
Syllabus
Intro
Can we Learn on the Edge? Al systems need to continually adapt to new data collected from the sensors Not only inference, but also run back-propagation on edge devices
#Activation is the Memory Bottleneck, not #Trainable Parameters
Related Work: Parameter-Efficient Transfer Learning
Address Optimization Difficulty of Quantized Graphs
QAS: Quantization-Aware Scaling
Sparse Layer/Tensor Update
Find Layers to Update by Contribution Analysis
Tiny Training Engine (TTE)
Tiny Training Engine Workflow
Deep Gradient Compression: Reduce Bandwidth
Post-training testing (high accuracy)
Real-life testing
Taught by
tinyML
Related Courses
-
On-Device Training Under 256KB Memory
-
On-Device Training and Transfer Learning - Lecture 15
-
On-Device Training and Transfer Learning - Lecture 15
-
On-device Training and Transfer Learning - Lecture 21
-
On-device Training and Transfer Learning in Machine Learning - Lecture 21
-
Getting Started with Machine Learning at the Edge on Arm
