Top-p Sampling
Sampling only from the smallest set of tokens whose probabilities reach p.
Visão geral
Top-p, or nucleus sampling, keeps the most likely tokens until their cumulative probability reaches a threshold, then samples from that set. It solves the problem that a fixed top-k includes nonsense when the distribution is peaked and excludes reasonable options when it is flat. It exists because the set of plausible continuations should be sized by the model's own confidence.
Grafo de tecnologias
Como Top-p Sampling se liga ao resto do stack.
Alternativas
-
Beam Search
technique
Uma alternativa a Top-p Sampling · source
Comparações
Comparações detalhadas com as alternativas acima.