Finetuned language model based on GPT-2 to generate Tweets in a users style. Tweety Face is a transformer model finetuned using GTP-2 and Tweets from various Twitter users. It was created to generate a Twitter Tweet for a given user similar to their specific writing style. It accepts a prompt for a user and completes the text. This finetuned model uses the smallest version of GPT-2, with 124M parameters. This model was created to experiment with prompt inputs and is not intended to create real Tweets. The generated text is not a real representation of the given users opinion, political affiliation, behaviour, etc. Do not use this model to impersonate a user. You can use this model directly with a pipeline for text generation. Since the generation relies on some randomness, we set a seed for reproducibility: The training data used for this model has been released as a dataset one can browse here. The raw data can be found in our Github Repository. The raw data can be found in two versions. All data on the develop branch is used in a debugging dataset. All data in the qa branch is used in the final dataset. For training first all retweets (RT) have been removed. Next the newline…
Модальности:
Генерация текста
Задача: Генерация текста
Автор: ML-Projects-Kiel
Теги: gpt2, en, text-generation-inference, endpoints_compatible
Лайков: 2 | Загрузок: 14
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.