llama.cpp
c252e0c4 - metal : optimize multi-sequence FA vec kernel (#13493)

Commit

211 days ago

metal : optimize multi-sequence FA vec kernel (#13493) * batched-bench : fix pp batch contents * metal : optimize multi-sequence FA vec kernel ggml-ci

References

#13493 - metal : optimize multi-sequence FA vec kernel

Author

ggerganov

Parents

4f711afe

llama.cpp c252e0c4 - metal : optimize multi-sequence FA vec kernel (#13493)

llama.cpp
c252e0c4 - metal : optimize multi-sequence FA vec kernel (#13493)