Saltar al contenido

Transformer

The neural network architecture behind modern LLMs, using self-attention to process all tokens in parallel.

Recursos relacionados