Metadata-Version: 2.1
Name: torchxlstm
Version: 0.0.2
Summary: A pure pytorch implementation of xLSTM.
Author-email: Akaash Dash <akaash.dash@gmail.com>
Maintainer-email: Akaash Dash <akaash.dash@gmail.com>
Project-URL: Homepage, https://github.com/akaashdash/xlstm
Project-URL: Source, https://github.com/akaashdash/xlstm/
Keywords: xlstm,lstm,torch,pytorch
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: torch

# xLSTM

A pure pytorch implementation of the [XLSTM paper](https://arxiv.org/abs/2405.04517).

## TODO
- Ensure correct with paper
- Create some usage examples
- Implement paralleization
- CUDA?
- Allow for different initializations according to: https://pytorch.org/docs/stable/nn.init.html
- Allow for flattening of x for greater shape conformity
- Allow for batching
- Add tests
    - https://github.com/catid/audio_prediction/tree/master
    - other classic RNN/LSTM tasks

## References
- https://arxiv.org/abs/2405.04517
- https://discuss.pytorch.org/t/causal-convolution/3456/3
- https://pytorch.org/docs/stable/generated/torch.nn.RNN.html
- https://pytorch.org/docs/stable/generated/torch.nn.LSTM.html
