Video The Transformer
Last updated: January 4, 2023
Jay Alammar is a blogger whose blogs have been referenced in MIT and Stanford courses due to their quality.
His blog on The Transformer in particular does a great job at explaining the most powerful deep learning model to date which gave rise to the Bidirectional Encoder Representations from Transformers (BERT) from Google and GPT-3 from OpenAI. Jay goes over that blog in the following 30 min video.
Transformers are extremely complex and this video is the best resource I have found to ease into understanding them.
For reference, here is the actual paper describing the Transformer.