Skip to content

Releases: l3utterfly/llama.cpp

b5061

07 Apr 10:07
916c83b

Choose a tag to compare

musa: fix compilation warnings in mp_22/31 (#12780)

Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>

b4959

26 Mar 05:23
53af4db

Choose a tag to compare

convert: fix Mistral3/Gemma3 model hparams init (#12571)

* Fix Mistral3/Gemma3 model hparams init

* set positional args correctly

* use existing hparams if passed

b4913

18 Mar 11:32
35cae5b

Choose a tag to compare

SYCL: using graphs is configurable by environment variable and compil…

b4879

13 Mar 10:25
f08f4b3

Choose a tag to compare

Update build.yml for Windows Vulkan builder to use Vulkan 1.4.304 SDK…

b4767

24 Feb 16:29
08d5986

Choose a tag to compare

[SYCL] Optimize mul_mat for Q4_0 on Intel GPU (#12035)

* opt performance by reorder for Intel GPU

* detect hw type and save opt feature, and print opt feature

* correct name

* support optimize graph once when compute graph, record the opt status in tensor->extra, make CI passed

* add env variable GGML_SYCL_DISABLE_OPT for debug

* use syclex::architecture replace the custom hw define, update the guide for GGML_SYCL_DISABLE_OPT

* add performance data

* mv getrows functions to separeted files

* fix global variables

---------

Co-authored-by: arthw <14088817+arthw@users.noreply.github.com>

b4713

14 Feb 08:38
a4f011e

Choose a tag to compare

vulkan: linux builds + small subgroup size fixes (#11767)

* mm subgroup size

* upload vulkan x86 builds

b4687

11 Feb 07:22
b9ab0a4

Choose a tag to compare

CUDA: use arch list for compatibility check (#11775)

* CUDA: use arch list for feature availability check

---------

Co-authored-by: Diego Devesa <slarengh@gmail.com>

b4519

21 Jan 05:58
80d0d6b

Choose a tag to compare

common : add -hfd option for the draft model (#11318)

* common : add -hfd option for the draft model

* cont : fix env var

* cont : more fixes

b4393

28 Dec 08:11
d79d8f3

Choose a tag to compare

vulkan: multi-row k quants (#10846)

* multi row k quant shaders!

* better row selection

* more row choices

* readjust row selection

* rm_kq=2 by default

b4302

11 Dec 07:41
43041d2

Choose a tag to compare

ggml: load all backends from a user-provided search path (#10699)

* feat: load all backends from a user-provided search path

* fix: Windows search path

* refactor: rename `ggml_backend_load_all_in_search_path` to `ggml_backend_load_all_from_path`

* refactor: rename `search_path` to `dir_path`

* fix: change `NULL` to `nullptr`

Co-authored-by: Diego Devesa <slarengh@gmail.com>

* fix: change `NULL` to `nullptr`

---------

Co-authored-by: Diego Devesa <slarengh@gmail.com>