llama.cpp
server: allow accessing /metrics and /slots during llama_decode()
#27041
Merged

Commits
  • server_queue::worker
    ngxson committed 2 days ago
  • call llama_decode inside yield_to_queue
    ngxson committed 2 days ago
  • also handle process_mtmd_chunk
    ngxson committed 2 days ago
  • clean up
    ngxson committed 2 days ago
  • nits
    ngxson committed 2 days ago
  • rm test
    ngxson committed 1 day ago
Loading