llama_cpp

Files

T

nullnameandGitHub 8af1f5f430 ggml-hexagon: flash-attn opt (#19025 )

* optimize flash attention kernel by improving score computation and online softmax update

* wip

* Refactor online softmax update in flash attention kernel for improved performance

* Optimize flash attention kernel by replacing float array with HVX_Vector for score computation

* wip

2026-01-23 22:02:07 -08:00

htp

ggml-hexagon: flash-attn opt (#19025 )

2026-01-23 22:02:07 -08:00

CMakeLists.txt

ggml-hexagon: Implement true Q8_0 quantization on Hexagon NPU for more accurate mixed-precision matmul operations (#17977 )

2025-12-19 09:42:28 -08:00

ggml-hexagon.cpp

ggml : add ggml_build_forward_select (#18550 )

2026-01-19 20:03:19 +02:00

htp-utils.c

ggml-hexagon: Initial Hexagon v68/v69 support (#17394 )