Decode
AIDecode: The generation stage of language model inference, where output tokens are produced one at a time after the prompt is processed.
Decode: The generation stage of language model inference, where output tokens are produced one at a time after the prompt is processed.