Definition
In deep learning, a transformer is a neural network architecture based on the self-attention mechanism, which weighs the relative relevance of all the elements of a sequence in order to process them in parallel. This allows it to capture long-distance dependencies without recurrence and forms the basis of today's language and multimodal models.
Relationships
Concepts linked to this one by a real relationship in the glossary. Hover over each one to see why it's related.