Parameters vs Tokens

Tokens are input or output chunks of text, while parameters are the learned weights inside a model. For example, GPT-4 has trillions of parameters and processes thousands of tokens per prompt.