What is Pre-training?
The process of training a foundation model on a massive dataset of general-purpose information before fine-tuning it for specific tasks. This allows the model to learn generalizable knowledge and representations that can be adapted to various domains and applications.