Named Entity Recognition API

An API for using the Estonian named entity recognition (NER) service. The service is based on the EstBERT_NER model.

Docker setup

The API can be deployed using the docker image published alongside the repository. Each image version correlates to a specific release. The application can be run with the following command:

docker run -p 8000:8000 ghcr.io/tartunlp/bert-ner-api:latest

The service is available on port 8000. The API documentation is available under the /docs endpoint.

Endpoints for healthcheck probes:

• /health/startup
• /health/readiness
• /health/liveness

Configuration

The following environment variables can be specified when running the container:

• API_MAX_INPUT_LENGTH (optional) - Maximum request size in character (10000 by default)
• API_CONFIG_PATH (optional) - Path to the configuration file (config/config.yaml by default)

Build-time arguments / environment variables:

• API_VERSION - A semantic version number, displayed in the docs.

The entrypoint of the container is ["uvicorn", "app:app", "--host", "0.0.0.0", "--proxy-headers", "--log-config", "config/logging.prod.ini"]. CMD can be used to define additional Uvicorn parameters. For example, ["--log-config", "config/logging.debug.ini", "--root-path", "/api/ner"] enables debug logging (as the last --log-config flag is used) and allows the API to be deployed to the non-root path /api/ner.

Manual setup

For a manual/development setup, the following steps can be used:

1. Clone this repository
2. Install the dependencies with pip install -r requirements.txt
3. Run the application with uvicorn app:app. The model files will be downloaded automatically upon startup.

Citation

If you use the model or this code in your research, please cite the following paper:

@misc{tanvir2020estbert,
      title={EstBERT: A Pretrained Language-Specific BERT for Estonian}, 
      author={Hasan Tanvir and Claudia Kittask and Kairit Sirts},
      year={2020},
      eprint={2011.04784},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}