Generative refers to the ability to generate new content, in this case text, based on a given input or prompt. In the context of the GPT language model, it means the ability to generate text that is coherent and contextually appropriate based on the input it has been trained on.
Pretrained refers to a machine learning model that has already been trained on a large corpus of data prior to being used for a specific task. The idea is that by using a pretrained model, the training time for a new task can be reduced significantly, and the model can perform better with less data.
The Transformer is a deep learning architecture used in Natural Language Processing (NLP) tasks, such as text classification, machine translation, and question-answering.