原书出处
Designing Large Language Model Applications
Beam Search An alternative to greedy decoding is beam search. An important parameter of beam search is the beam size, n. At the first step, the top n tokens with the highest probabilities are selected as hypotheses. For the next few steps, the model generates token continuations for each of the hypotheses. The token chosen to be generated is the one whose continuations have the
原书定位详情
来源:a19c7933-4eda-4176-aebd-30fe7e1ad292
正文位置:extracted/Designing Large Language Model Applications/_epub-src/OEBPS/ch05.html
EPUB SHA-256:1b4d2347fec2f949fcddfbaf4fad597a1ae20887a096cfae64c59a619b4781bf