Transformers are neural networks used in the field of sequence processing, such as text or speech.
Unlike earlier models like RNNs or LSTMs, Transformers process all input elements simultaneously and use the so-called attention mechanism.
This mechanism allows the model to evaluate the importance of different parts of the input and combine relevant information from different positions.
Typically, transformer models consist of multiple layers of self-attention and feed-forward networks that together produce deep representations.
Well-known applications are language models like BERT or GPT, which are based on transformer architectures.