Replay planned for dotAI talk on llama.cpp speculative decoding
The speaker says the talk covers MTP, dflash and dspark.
TLDR
A dotAI speaker announced a forthcoming replay of their talk on speculative decoding in llama.cpp, covering MTP, dflash and dspark. On September 17, 2026, they said the replay would be available soon.
Combined views
115
1 Source, first seen ago