Gpt2 Base Style Paraphraser
Unsupervised style transfer model based on GPT2 architecture, achieving text style conversion through paraphrase generation
Downloads 64
Release Time : 3/2/2022
Model Overview
This model originates from the research by Krishna et al., reframing unsupervised style transfer as a paraphrase generation task, capable of altering writing styles while preserving the original text semantics
Model Features
Unsupervised Style Transfer
Achieves text style conversion without requiring parallel corpora
Paraphrase Generation Framework
Reframes style transfer as a paraphrase generation task
Semantic Preservation
Maintains original text semantics while transforming styles
Model Capabilities
Text Style Transfer
Paraphrase Generation
Unsupervised Learning
Use Cases
Text Editing
Formalization Rewriting
Converts informal text into formal style
Changes tone and word choice while preserving meaning
Style Adaptation
Adjusts text style according to target audience
Makes content more suitable for specific readers
Content Creation
Diverse Expression
Generates different style versions for the same content
Enhances content diversity and adaptability
Featured Recommended AI Models
Š 2025AIbase