Skip to content
Snippets Groups Projects
user avatar
Katrien Beuls authored
e3de37f1
History

text-to-role-classification

Webserver hosting a semantic role labeling model. Can make predictions

  • argm-phrases.json The labeled training data
  • train_model.py Trains the model from the file argm-phrases.json and stores the model in model.pkl
  • model.pkl The model to be used in the webserver to make predictions
  • server.py The flask webserver hosting the model for predictions
  • example_request.py Example of using the API.

Usage:

  • installation:
    • pip install -r requirements.txt
  • training:
    • python train_model.py
  • run API server:
    • python server.py
  • API usage:
    • POST
    • http://localhost:5000/predict
    • jsondata = ["Great", "Bad"]
    • curl -H "Content-Type: application/json" -X POST -d '["Great", "Bad"]' http://localhost:5000/predict