Publication

Structured and tiled-based pruning of Deep Learning models targeting FPGA implementations

Abstract

Model compression techniques have lead to a reduction of size and number of computations of Deep Learning models. However, techniques such as pruning mostly lack of a real co-optimization with hardware platforms. For instance, implementing unstructured pruning in dedicated hardware is not a straightforward task, which increases memory and reduces the effective bandwidth usage. Moreover, such pruning algorithms should be adapted to certain hardware requirements, such as the use of tiling. Therefore, in this work, we leverage the use of the Gumbel-Softmax relaxation sampling to structurally prune tiles, which benefits further hardware implementations, and additionally allows to jointly optimize with quantization. Additionally, we show that the combination of different pruning scenarios leads to a larger sparsity. Finally, we demonstrate the benefit of using structured pruning on fine-grained elements (weights) in an FPGA design.

About this result
This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.
Related concepts (34)
Hardware virtualization
Hardware virtualization is the virtualization of computers as complete hardware platforms, certain logical abstractions of their componentry, or only the functionality required to run various operating systems. Virtualization hides the physical characteristics of a computing platform from the users, presenting instead an abstract computing platform. At its origins, the software that controlled virtualization was called a "control program", but the terms "hypervisor" or "virtual machine monitor" became preferred over time.
Open-source hardware
Open-source hardware (OSH) consists of physical artifacts of technology designed and offered by the open-design movement. Both free and open-source software (FOSS) and open-source hardware are created by this open-source culture movement and apply a like concept to a variety of components. It is sometimes, thus, referred to as FOSH (free and open-source hardware). The term usually means that information about the hardware is easily discerned so that others can make it – coupling it closely to the maker movement.
Data compression
In information theory, data compression, source coding, or bit-rate reduction is the process of encoding information using fewer bits than the original representation. Any particular compression is either lossy or lossless. Lossless compression reduces bits by identifying and eliminating statistical redundancy. No information is lost in lossless compression. Lossy compression reduces bits by removing unnecessary or less important information.
Show more
Related publications (38)

Graph Chatbot

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.