Commit History

Автор SHA1 Съобщение Дата
  AlpinDale d8a51d05a7 fix: seeded gens with pipeline parallel преди 5 месеца
  AlpinDale 2c653a2268 fix: make speculative decoding work with per-request seed преди 5 месеца
  AlpinDale af43576da0 feat: add MLPSpeculator speculative decoding support (#572) преди 6 месеца
  AlpinDale 4d1e613804 chore: minor simplifications преди 6 месеца
  AlpinDale e0886ee929 feat: add `ProposerWorkerBase` abstract class преди 6 месеца
  AlpinDale a94de94c44 refactor: combine the prefill and decode into a single API (#553) преди 6 месеца
  AlpinDale ef733aee43 implement ExecuteModelData to reduce executor complexity преди 6 месеца
  AlpinDale 79901b76de logprobs for target model (spec decoding) преди 6 месеца
  AlpinDale 723c6acb84 re-add ngram speculative decoding преди 6 месеца
  AlpinDale 9d81716bfd [v0.5.3] Release Candidate (#388) преди 9 месеца
  AlpinDale f8dfac6372 chore: attention refactor and upstream sync apr01 (#365) преди 10 месеца