Skip to content

Source code for the web backend for hosting story generation models in the EMNLP 2020 paper "STORIUM: A Dataset and Evaluation Platform for Human-in-the-Loop Story Generation"

License

Notifications You must be signed in to change notification settings

dojoteef/storium-backend

Repository files navigation

Storium Backend Web Service

This is the official repository of the backend web service for the EMNLP 2020 long paper STORIUM: A Dataset and Evaluation Platform for Machine-in-the-Loop Story Generation.

Python Package Naming

You might ask yourself, what is a figmentator? A figmentator generates figments of the imagination. In the context of this code, it represents a story generation model. Why the weird name? Well it's distinctive and I was already woolgathering...

Usage

There are are currently three environments: dev, stage, & prod. Run make build-dev or make build-prod to create the docker containers to deploy the service.

Afterward you can run deploy-dev or make deploy-prod to start the service.

Deployment

Our evaluation platform is designed to make it easy to deploy your story generation models, requiring the implementation of a simple API to facilitate hosting. You can see a contrived example in src/figmentator/examples/simple.py, or a real example from our GPT-2 models.

Cite

@inproceedings{akoury2020storium,
  Author = {Nader Akoury, Shufan Wang, Josh Whiting, Stephen Hood, Nanyun Peng and Mohit Iyyer},
  Booktitle = {Empirical Methods for Natural Language Processing},
  Year = "2020",
  Title = {{STORIUM}: {A} {D}ataset and {E}valuation {P}latform for {S}tory {G}eneration}
}

About

Source code for the web backend for hosting story generation models in the EMNLP 2020 paper "STORIUM: A Dataset and Evaluation Platform for Human-in-the-Loop Story Generation"

Topics

Resources

License

Stars

Watchers

Forks