spacy_ner_conll2003_en

The project contain the following notebooks:

• parse_data.ipynb - heavy notebook, it's results are stored in data directory, so that you can start from the explore_conll2003_en.ipynb
• explore_conll2003_en.ipynb - lighweight notebook with data exploration
• train_ner_spacy_conll2003.ipynb - can train on CPU in reasonable time

Steps to create docker container and run it (ner_spacy_ns is a container name, you can change it):

• cd ner_proj/docker_ns
• docker build -t ner_spacy_ns .
• docker run -d -p 5000:5000 -it ner_spacy_ns

To run inference example (or same url from browser):

• curl http://0.0.0.0:5000/example

To run NER on your text from the command line

• place you text after the =, like in example below. curl http://0.0.0.0:5000/inp_text=Daniel%20is%20travelling%20to%20US

To run from browser

• open a browser and go to url: http://0.0.0.0:5000/inp_text=you text without ecape characters

If you need to stop and remove other containers that take the ports above:

• docker ps
• docker stop container_id
• docker rm container_id

If there any problem and you'd like to view cintainer logs

• note: container_id is taken from docker ps
• docker logs container_id