HASZB_AIHASZB_AI

Search across courses, lessons, glossary terms, prompts and tools.

models

Transformer

In one sentence

The neural network architecture behind modern language models.

In more depth

Introduced in 2017, its central mechanism is attention: when processing each token, the model weighs how much every other token in the context matters to it.

This let models handle long-range relationships in text and — critically — train efficiently on parallel hardware, which is what made training at today's scale practical.