PrithivirajDamodaran / Parrot_Paraphraser

A practical and feature-rich paraphrasing framework to augment human intents in text form to build robust NLU models for conversational engines. Created by Prithiviraj Damodaran. Open to pull requests and other forms of collaboration.
Apache License 2.0
866 stars 141 forks source link

Make it work without removing HTML tags #35

Closed Emilianocm23 closed 1 year ago

Emilianocm23 commented 1 year ago

Hello everyone, im using this paraphraser model and it works well. But i would like to passphrase my text without breaking html tags.

Example of the original text:

In Java, an interface specifies the behavior of a class by providing an abstract type. As one of Java's core concepts, abstraction, polymorphism, and multiple inheritance are supported through this technology. Interfaces are used in Java <b>to achieve abstraction</b>.

Example of the paraphrased text:

An interface in Java gives the behavior of a class. One of the core concepts of Java is the use of abstraction, polymorphism, and multiple inheritance. Java uses interface to achieve abstraction.

As you can see its removing the HTML tag, sometimes it doesn't remove it but break it them like this: "/li>li>"

Any help is appreciated, i can pay if you can solve this for me