Repository navigation

[TRTLLM-17267][refactor] Move the embeddings, mm_encoder and tokenize handlers into serving classes - #21575

Draft

JunyiXu-nv wants to merge 31 commits into

NVIDIA:mainfrom

JunyiXu-nv:dev-junyix-feat-serve-embeddings-mm-encoder

Draft

[TRTLLM-17267][refactor] Move the embeddings, mm_encoder and tokenize handlers into serving classes#21575
JunyiXu-nv wants to merge 31 commits into
NVIDIA:mainfrom
JunyiXu-nv:dev-junyix-feat-serve-embeddings-mm-encoder

Commits on Nov 1, 2026
- [TRTLLM-17272][feat] Add public LLM health queries and a private-access lint for the serving frontends
  JunyiXu-nv
  committed
  259fdf6
- Fix frontend lint coverage and health test cleanup
  JunyiXu-nv
  committed
  ef64f31
Commits on Nov 2, 2026
- [TRTLLM-17272][feat] Add public runtime accessors for the remaining serving-frontend private accesses
  JunyiXu-nv
  committed
  73c213b
- [TRTLLM-17272][test] Call _on_launcher_lost on a stand-in instead of a bare OpenAIServer
  JunyiXu-nv
  committed
  d4b6d14
Commits on Nov 4, 2026