Skip to content

Releases: ngxson/llama.cpp

b5516

27 May 20:28
a3c3084
Compare
Choose a tag to compare
opencl: add new ops - `argsort`, `div`, `sub`, `addrows`, `sigmoid`, …

b5515

27 May 20:23
1701d4c
Compare
Choose a tag to compare
opencl: mark `mul_mat` `f32f32` as supporting non-contiguous tensors …

b5514

27 May 17:13
bef8176
Compare
Choose a tag to compare
vulkan: use timestamp queries for GGML_VULKAN_PERF (#13817)

Also change it to be controlled by an env var rather than cmake flag

b5513

27 May 16:36
34b7c04
Compare
Choose a tag to compare
cmake : add llama-cparams.cpp to build (#13832)

b5512

27 May 15:50
f3101a8
Compare
Choose a tag to compare
SYCL: add gelu_erf kernel (#13749)

* SYCL: add gelu_erf kernel

* refactor code

Co-authored-by: Atharva Dubey <[email protected]>

* Use scope_op_debug_print

---------

Co-authored-by: Atharva Dubey <[email protected]>

b5510

27 May 15:01
a8ea03d
Compare
Choose a tag to compare
ggml : add ggml_repeat_4d (#13824)

b5509

27 May 14:02
05f6ac6
Compare
Choose a tag to compare
ggml : riscv: add xtheadvector support (#13720)

* ggml : riscv: add xtheadvector support

* ggml : clean up some macro usage

b5508

27 May 12:23
bc583e3
Compare
Choose a tag to compare
mtmd : support Qwen 2.5 Omni (input audio+vision, no audio output) (#…

b5506

27 May 11:43
7fe03e7
Compare
Choose a tag to compare
ggml-cpu: x86 feature detection is specific to x86 (#13811)

b5505

27 May 11:35
952f395
Compare
Choose a tag to compare
ggml : allow CUDA graphs when using pipeline parallelism (#13814)