misc: make max_top_p/k_rounds a input argument instead of template parameter - Githubissues

flashinfer-ai / flashinfer

FlashInfer: Kernel Library for LLM Serving

https://flashinfer.ai

Apache License 2.0

822 stars 77 forks source link

misc: make max_top_p/k_rounds a input argument instead of template parameter #219

Closed yzh119 closed 2 months ago

yzh119 commented 2 months ago

max_top_p/k_rounds doesn't have to be a template parameter.