Commit History

Autor SHA1 Mensaxe Data
  AlpinDale 3a53ff1e01 fix: raise an error for no draft token case when draft_tp>1 hai 5 meses
  AlpinDale d638dc592d fix: some minor typing issues in spec decode hai 5 meses
  AlpinDale 16dff9babc chore: enable bonus token in spec decoding for KV cache based models hai 5 meses
  AlpinDale af43576da0 feat: add MLPSpeculator speculative decoding support (#572) hai 5 meses
  AlpinDale 4d1e613804 chore: minor simplifications hai 5 meses
  AlpinDale e0886ee929 feat: add `ProposerWorkerBase` abstract class hai 6 meses
  AlpinDale e42d0b3455 possibly improve ngram efficiency hai 6 meses
  AlpinDale 197a6d2c16 auto disable speculative decoding by the running queue size hai 6 meses
  AlpinDale ef733aee43 implement ExecuteModelData to reduce executor complexity hai 6 meses
  AlpinDale 79901b76de logprobs for target model (spec decoding) hai 6 meses
  AlpinDale 9d81716bfd [v0.5.3] Release Candidate (#388) hai 8 meses