I meant deep neural networks with transformer architecture, and self-attention so they can be trained using GPUs. Doesn't have to be specifically "large languag

I meant deep neural networks with transformer architecture, and self-attention so they can be trained using GPUs. Doesn't have to be specifically "large language" models necessarily, if that's your hangup.