Metadata-Version: 2.1
Name: LRU-tensorflow
Version: 0.1.1
Summary: Unofficial TensorFlow implementation of a Linear Recurrent Unit, proposed by Google Deepmind.
Home-page: https://github.com/DemandredEng/Linear-Recurrent-Units-Tensorflow
Download-URL: 
Author: Udit Sharma
Author-email: uditsharma.eng@gmail.com
License: MIT
Keywords: Artificial Intelligence,Deep Learning,Recurrent Neural Networks,Linear Recurrent Unit
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Description-Content-Type: text/markdown
License-File: LICENSE.txt

# Linear Recurrent Units in Tensorflow: An Unofficial Implementation
This repository presents an unofficial implementation of Linear Recurrent Units (LRUs) proposed by Google DeepMind, utilizing Tensorflow. LRUs draw inspiration from Deep State-Space Machines, with a particular focus on S4 and S5 models.

# Installation:
```
$ pip install LRU-tensorflow
```
# Usage:
```python
import tensorflow as tf
from LRU_tensorflow import LRU

lru = LRU(N=state_features, H=input_size) 
test_input = tf.random.uniform((batch_size, seq_length, input_size))  # Example Test Input
predictions = lru(test_input) # Get predictions
```

# Paper:
<a href='https://arxiv.org/abs/2303.06349'>Resurrecting Recurrent Neural Networks for Long Sequences</a>


# Notes:
+ If you require an implementation that supports 3-dimensional input sequences, you may want to refer to <a href='https://github.com/Gothos/LRU-pytorch'>github.com/Gothos/LRU-pytorch</a>. However, please be aware that this alternative implementation might be slower due to the absence of associative scans.
