The impressive capability of LLMs to comprehend and draw inferences from context and generate human-like relevant responses for diverse tasks such as translation, summarization, question answering, and creative endeavors like poetry or code generation has left people amazed worldwide.
[...]
Language models are pre-trained to predict the next token in a text corpus. How to choose the next token based on the probability distribution over a fixed vocabulary, which is where decoding strategies come into play.
