Objective: Demonstrate ability to research a new NLP concept and create a proof of concept for evaluation
Task Description:
- Research “zero shot classification” pipeline feature of the HuggingFace transformers NLP library for python
- Incorporate the “zero shot classification” pipeline code in a backend Flask app with one HTTP POST endpoint
- The HTTP POST endpoint should expect a JSON payload that has two values: a string sentence and an array of categories
- The HTTP POST endpoint should process the input data using the HuggingFace library and then return a JSON payload that indicates the percent probability that the sentence belongs to each of the categories in the input JSON payload
- Test the Flask backend app on your local machine
- Upload the tested code to a GitHub repo in your account (create a new account if needed)
- Email the GitHub repo link to careers at seealgo dot com