guest ::
| Home > Publications database > Pre-training and Meta-learning for Memristor Crossbar Arrays |
| Home > Publications database > Pre-training and Meta-learning for Memristor Crossbar Arrays |
| Conference Presentation (Invited) | FZJ-2024-02777 |
2023
This record in other databases:
Please use a persistent id in citations: doi:10.29363/nanoge.neuronics.2024.027
Abstract: Memristive crossbar arrays show promise as non-von Neumann computing technologies, bringing sophisticated neural network processing to the edge and facilitating real-world online learning. However, their deployment for real-world learning problems faces challenges such as non-linearities in conductance updates, variations during operation, fabrication mismatch, conductance drift, and the realities of gradient descent training.This talk will present methods to pre-train neural networks to be largely insensitive to these non-idealities during learning tasks. These methods rely on a phenomenological model of the device, obtainable experimentally, and bi-level optimization. We showcase this effect through meta-learning and a differentiable model of conductance updates on few-shot learning tasks. Since pre-training is a necessary procedure for any online learning scenario at the edge, our results may pave the way for real-world applications of memristive devices without significant adaptation overhead.Furthermore, by considering the programming of memristive devices as a learning problem in its own right, we demonstrate that the developed methods can accelerate existing write-verify techniques.
|
The record appears in these collections: |
