light-recurrent-unit-pytorch 0.0.4

Light Recurrent Unit

Light Recurrent Unit - Pytorch (wip)

Implementation of the Light Recurrent Unit in Pytorch

Install

$ pip install light-recurrent-unit-pytorch

Usage

import torch
from light_recurrent_unit_pytorch import LightRecurrentUnitCell

lru_cell = LightRecurrentUnitCell(256)

x = torch.randn(2, 256)
hidden = torch.randn(2, 256)

next_hidden = lru_cell(x, hidden) # (2, 256)

Single layer

import torch
from light_recurrent_unit_pytorch import LightRecurrentUnitLayer

lru_layer = LightRecurrentUnitLayer(256)

x = torch.randn(2, 1024, 256)

output = lru_layer(x) # (2, 1024, 256)
assert x.shape == output.shape

Stacked

import torch
from light_recurrent_unit_pytorch import LightRecurrentUnit

lru = LightRecurrentUnit(256, depth = 4)

x = torch.randn(2, 1024, 256)

out = lru(x)

assert out.shape == x.shape

Citations

@Article{electronics13163204,
    AUTHOR = {Ye, Hong and Zhang, Yibing and Liu, Huizhou and Li, Xuannong and Chang, Jiaming and Zheng, Hui},
    TITLE = {Light Recurrent Unit: Towards an Interpretable Recurrent Neural Network for Modeling Long-Range Dependency},
    JOURNAL = {Electronics},
    VOLUME = {13},
    YEAR = {2024},
    NUMBER = {16},
    ARTICLE-NUMBER = {3204},
    URL = {https://www.mdpi.com/2079-9292/13/16/3204},
    ISSN = {2079-9292},
    DOI = {10.3390/electronics13163204}
}