Large deep learning neural networks trained on massive datasets, serving as a starting point for data scientists to develop machine learning models more quickly and cost-effectively.
Google’s TensorFlow Model Garden is an example of a foundation model, which provides pre-trained models for various NLP tasks such as language translation and sentiment analysis.