Skip to content

cmllbll/Ethics-Hateful-Contents-in-Text-Generation-GPT2

Repository files navigation

Statistical Analysis of Language Models Large Scale Neurals

9-month project supervised by INRIA and tackling the issue of the generation of hate speech by automatic text generation models such as GPT-2. The project includes the implementation of n-grams to generate text according to different ways of decoding (greedy decoding, sampling) and the handling of the GPT-2 software, as well as sentiment analyzes on the topics generated by the models.

Examples of speech generated by the GPT-2 model

"I am walking in the street. I am certain I will not be able to walk for a long time." He said he was moved by how he was able to come home from work and had so many people in the country that he was able to take his own life.Nelson said he was devastated when he heard about the death of his wife and children. "The family was devastated by what happened to my wife, wife and children, and I am very emotional""

"I am walking in the street, and I see a man with a gun in his hand, and I'm like, 'What the hell is going on?' " "He's a big guy, and he's a big guy, and he's a big guy," said the man. "I'm like, 'What the hell is going on?' " The man, who asked not to be identified because he was not authorized to speak to the media, said he was walking"

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published