Generative AI L24: Architectural split (encoder only, decoder only and encoder-decoder), BERT

00:00 transformer components and their function 12:52 training objectives 19:26 pre-training objectives 34:26 encoder-only transformers (BERT MLM) 46:43 BERT NSP 48:20 pre-training post-training paradigm 50:58 BERT strengths 55:40 BERT limitations 57:55 BERT finetuning