llama.cpp
ggml-cpu: aarm64: q4_K repack gemm and gemv implementations (dotprod only)
#17494
Merged
Overview
Commits
14
Changes
View On GitHub
Commits
  • Enabled q4_K_4x8 path
    Alcpz committed 36 days ago
  • Fixed generic Q4_K 8x4 implementation
    Alcpz committed 36 days ago
  • wip: dotprod gemm
    Alcpz committed 36 days ago
  • Working arm q4_K dotprod gemm
    Alcpz committed 36 days ago
  • Undo acc rename
    Alcpz committed 36 days ago
  • Q4_K arm dotprod gemm
    Alcpz committed 36 days ago
  • Fix: q4_qs reinterpret from uint to int
    Alcpz committed 36 days ago
  • Removed comments
    Alcpz committed 36 days ago
  • Fixed macro guards
    Alcpz committed 36 days ago
  • Fixed unused vars in generic implementation
    Alcpz committed 36 days ago
  • Fixed unused vars in 8x4 repack
    Alcpz committed 36 days ago
  • Fixed unused vars in generic implementation, unneeded comment
    Alcpz committed 36 days ago
  • Missing arch fallback for x86
    Alcpz committed 36 days ago
  • minor : style
    ggerganov committed 34 days ago
FAQ Terms Privacy Refunds Impressum
Loading