https://github.com/1ytic/pytorch-edit-distance

Levenshtein edit-distance on PyTorch and CUDA
https://github.com/1ytic/pytorch-edit-distance

asr cuda edit-distance levenshtein nlp pytorch

Last synced: about 7 hours ago
JSON representation

Levenshtein edit-distance on PyTorch and CUDA

• Host: GitHub
• URL: https://github.com/1ytic/pytorch-edit-distance
• Owner: 1ytic
• License: mit
• Created: 2019-11-02T22:16:50.000Z (about 5 years ago)
• Default Branch: master
• Last Pushed: 2023-01-24T21:06:51.000Z (about 2 years ago)
• Last Synced: 2025-01-22T18:08:52.258Z (8 days ago)
• Topics: asr, cuda, edit-distance, levenshtein, nlp, pytorch
• Language: Cuda
• Size: 9.77 KB
• Stars: 94
• Watchers: 8
• Forks: 14
• Open Issues: 4
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# PyTorch edit-distance functions

Useful functions for E2E Speech Recognition training with PyTorch and CUDA.

Here is a simple use case with Reinforcement Learning and RNN-T loss:

```python

blank = torch.tensor([0], dtype=torch.int).cuda()

space = torch.tensor([1], dtype=torch.int).cuda()

xs = model.greedy_decode(xs, sampled=True)

torch_edit_distance.remove_blank(xs, xn, blank)

rewards = 1 - torch_edit_distance.compute_wer(xs, ys, xn, yn, blank, space)

nll = rnnt_loss(zs, ys, xn, yn)

loss = nll * rewards

```

### levenshtein_distance

Levenshtein edit-distance with detailed statistics for ins/del/sub operations.

### collapse_repeated

Merge repeated tokens, useful for CTC-based model.

### remove_blank

Remove unnecessary blank tokens, useful for CTC, RNN-T, RNA models.

### strip_separator

Remove leading, trailing and repeated middle separators.

## Requirements

- C++11 compiler (tested with GCC 9.4.0).

- Python: 3.5, 3.6, 3.7, 3.8, 3.9 (tested with version 3.8).

- [PyTorch](http://pytorch.org/) >= 1.5.0 (tested with version 1.13.1+cu116).

- [CUDA Toolkit](https://developer.nvidia.com/cuda-zone) (tested with version 11.2).

## Install

There is no compiled version of the package. The following setup instructions compile the package from the source code locally.

### From Pypi

```bash

pip install torch_edit_distance

```

### From GitHub

```bash

git clone https://github.com/1ytic/pytorch-edit-distance

cd pytorch-edit-distance

python setup.py install

```

## Test

```bash

python -m torch_edit_distance.test

```