LSTM-based seq2seq number sort

JetBrains internship test assignment solution

pytorch baseline - autoencoder.ipynb

solution - LSTM Sorter.ipynb

possibly some mistake in lstm gradient, because model trains very slow in comparasion with torch implementation
typing incompatible with mypy

Name		Name	Last commit message	Last commit date
Latest commit History 14 Commits
LSTM Sorter.ipynb		LSTM Sorter.ipynb
README.md		README.md
autoencoder.ipynb		autoencoder.ipynb
dataset.py		dataset.py
layers.py		layers.py
loss.py		loss.py
metrics.py		metrics.py
model.py		model.py
module.py		module.py
operators.py		operators.py
optim.py		optim.py
trainer.py		trainer.py