README.md

metadata

widget:
  - text: 'generate analogy: mammal is to whale'
    example_title: Analogy Example 1 (semantic relation)
  - text: 'generate analogy: wedding is to marriage'
    example_title: Analogy Example 1 (semantic relation, metaphor)
  - text: 'generate analogy: London is to U.K.'
    example_title: Analogy Example 2 (entity)
  - text: 'generate analogy: actual is to actually'
    example_title: Analogy Example 3 (morphological)

relbert/flan-t5-xl-analogy

This is google/flan-t5-xl fine-tuned on relbert/semeval2012_relational_similarity for analogy generation, which is to generate a word pair (eg. bird is to crow) given a query (eg. mammal is to whale) so that the query and the generated word pair form an analogy statement.

Usage

from transformers import pipeline

pipe = pipeline('text2text-generation', model="relbert/flan-t5-xl-analogy")
output = pipe("generate analogy: mammal is to whale")
print(output)
>>> [{'generated_text': 'bird is to crow'}]