Skip to content

shashwat1198/FINN-GLSTM-Hw

Repository files navigation

FINN-GLSTM-hw

  • This repository contains the hardware implementation of a streamlined LSTM layer for use with FINN, leveraging finn-hlslib building blocks.
  • Paper link: https://arxiv.org/pdf/2506.20810

Overview

  • The top-level computation is defined in qlstm_top.cpp, which represents the operations derived from the streamlined LSTM QONNX graph.
  • Layer-specific parameters (size of inputs, outputs, lookbacks and datatypes) are provided in pipeline-lstm-header.h.

Usage

The generated hardware can be exported as a synthesizable IP block, which integrates seamlessly with other FINN-compatible layers such as convolutional and fully connected layers. This enables the extension of FINN to support Recurrent Neural Networks (RNNs), particularly LSTM-based architectures.

Key Features

  • Streamlined and synthesizable LSTM design
  • AXI-Stream interface for easy integration
  • Compatible with FINN-generated QONNX models
  • Modular IP for combining with other layers

About

QLSTM hardware layer constructed with finn-hlslib hardware blocks

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages