https://github.com/desh2608/streaming_wer

CUDA implementation of Streaming WER metric
https://github.com/desh2608/streaming_wer

Last synced: 2 months ago
JSON representation

CUDA implementation of Streaming WER metric

• Host: GitHub
• URL: https://github.com/desh2608/streaming_wer
• Owner: desh2608
• License: mit
• Created: 2023-07-27T14:00:54.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2023-07-27T15:17:44.000Z (almost 2 years ago)
• Last Synced: 2025-01-24T22:32:10.086Z (4 months ago)
• Language: Python
• Size: 13.7 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
Streaming WER


This is a CUDA implementation of the streaming word error rate metric (streaming WER),

proposed by Ehsan Variani during JSALT 2023.

## Streaming WER

## Installation

```bash

pip install -e .

```

Run tests:

```bash

pytest tests

```

## Usage

## Credits

This implementation borrows heavily from the following repository: https://github.com/1ytic/pytorch-edit-distance