README.md

metadata

widget:
  - text: >-
      context: I would hope that Phylicia Rashad would apologize now that
      @missjillscott has! You cannot discount 30 victims who come with similar
      stories.— JDWhitner (@JDWhitner) July 7, 2015, answer: apologize
    example_title: example 1
  - text: >-
      context: I would hope that Phylicia Rashad would apologize now that
      @missjillscott has! You cannot discount 30 victims who come with similar
      stories.— JDWhitner (@JDWhitner) July 7, 2015, answer: 30
    example_title: example 2
  - text: >-
      context: The news about Vegas is devastating. Sending all our love to the
      people there right now ❤️❤️❤️— HAIM (@HAIMtheband) October 2, 2017,
      answer: vegas
    example_title: example 3

cardiffnlp/flan-t5-small-tweet-qg

This is google/flan-t5-small fine-tuned on cardiffnlp/super_tweeteval (tweet_qg).

Usage

from transformers import pipeline

pipe = pipeline('text2text-generation', model="cardiffnlp/flan-t5-small-tweet-qg")
output = pipe("context: I would hope that Phylicia Rashad would apologize now that @missjillscott has! You cannot discount 30 victims who come with similar stories.— JDWhitner (@JDWhitner) July 7, 2015, answer: apologize")