upstream-archive/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-09-25 15:37:25 -05:00
Code Issues Packages Projects Releases Wiki Activity
Files
dbedc9e19c50dca0acdfb402362e2707bee424ae
llama.cpp/include
T
History
Yaniss 1f31996c50 speculative: distinguish draft heads from nextn block count 
A single MTP draft step can span multiple nextn blocks (NemotronH Puzzle:
[attention, moe]). Add llama_model_n_nextn_heads() and use it in draft-mtp
so a multi-block single-step head is not misread as chained heads.
2026-07-26 11:30:37 +02:00
..
llama-cpp.h
llama : re-enable manual LoRA adapter free (#19983)
2026-03-18 12:03:26 +02:00
llama.h
speculative: distinguish draft heads from nextn block count
2026-07-26 11:30:37 +02:00