https://github.com/tim-smart/label-studio-spacy
spaCy powered Label Studio ML backend
https://github.com/tim-smart/label-studio-spacy
Last synced: 9 months ago
JSON representation
spaCy powered Label Studio ML backend
- Host: GitHub
- URL: https://github.com/tim-smart/label-studio-spacy
- Owner: tim-smart
- License: mit
- Created: 2022-07-20T03:15:41.000Z (almost 4 years ago)
- Default Branch: main
- Last Pushed: 2022-12-11T20:09:10.000Z (over 3 years ago)
- Last Synced: 2025-06-07T16:48:19.837Z (about 1 year ago)
- Language: Python
- Size: 41 KB
- Stars: 29
- Watchers: 4
- Forks: 7
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# spaCy powered Label Studio ML backend
spaCy integration for Label Studio. Created as an Open Source alternative to Prodigy.
Benefits:
* Speed up annotation of data with integrated predictions
* Quickly iterate on your spaCy models
## Demo video
[](https://youtu.be/F19NT-21uT4)
## Usage
1. Clone this repo
2. Install requirements
```
pip install -r requirements.txt
```
3. Initialize a new backend
```
label-studio-ml init my_ml_backend
```
4. In the `my_ml_backend` directory, add your spaCy `config.cfg` file. You can optionally add a `model-best` folder from a pre-trained model, to get started with predictions straight away.
5. Start the backend and add the URL to your Label Studio project settings.
```
label-studio-ml start my_ml_backend
```
6. As you train new models, they will appear in a `checkpoints` directory. The latest checkpoint will be symlinked to `latest-model`.