llama.cpp
llama : fix DFlash output head sharing
#30111
Merged

Commits
  • llama : fix DFlash output head sharing
    hthadicherla committed 42 days ago
  • dflash : read tied output weights from GGUF metadata
    hthadicherla committed 42 days ago
  • llama : share tied word embedding metadata
    hthadicherla committed 41 days ago
  • llama : remove DFlash embedding head fallback
    hthadicherla committed 41 days ago
Loading