koboldcpp/scripts/snapdragon
Max Krasnyansky 9c2e0e491a
hexagon: enable alloc_buffer_n (#30126)
* hex-bufs: add support for alloc_buffer_n

* hex-bufs: add support for splitting large tensors into separate buffers

* hex-bufs: update GGML_HEXAGON_MBUF to accept three values dyn,static,total

* hex-bufs: bump dyn. default to 512MB since 128MB causes perf regressions with big MOEs

* hex-run: add --no-embd-offload option to simplify command lines on devices that need it

* Update scripts/snapdragon/run.py

Co-authored-by: Jhen-Jie Hong <iainst0409@gmail.com>

---------

Co-authored-by: Jhen-Jie Hong <iainst0409@gmail.com>
2026-10-07 18:30:06 -07:00
..
adb hexagon: support for multi-NPU devices (IQ9, IQ10) and fully asynchronous backend (#26501) 2026-08-26 18:46:50 -07:00
qdc snapdragon: ci updates to use new run script (#28293) 2026-09-03 08:59:11 -07:00
build.py snapdragon: ci updates to use new run script (#28293) 2026-09-03 08:59:11 -07:00
ggml-hexagon-align-macros.py hexagon: support for multi-device model split (aka row-split) (#28589) 2026-09-11 20:46:51 -07:00
ggml-hexagon-inspect.py hexagon: find software divide calls using binary inspection tool (#29449) 2026-09-25 19:43:38 -07:00
ggml-hexagon-profile.py hexagon: new HMX-optimized GATED_DELTA_NET (#29199) 2026-09-21 14:49:52 -07:00
ggml-hexagon-trace.py hex-scripts: show trace events smaller than 100nsec in perfetto (#29614) 2026-09-28 15:02:58 -07:00
run.py hexagon: enable alloc_buffer_n (#30126) 2026-10-07 18:30:06 -07:00
sdk.py snapdragon: python SDK setup (Windows) (#27903) 2026-08-28 14:01:59 -07:00
setup-sdk.py snapdragon: python SDK setup (Windows) (#27903) 2026-08-28 14:01:59 -07:00