Skip to content

Releases: ggml-org/llama.cpp

b5797

01 Jul 16:02
de56944
Compare
Choose a tag to compare
ci : disable fast-math for Metal GHA CI (#14478)

* ci : disable fast-math for Metal GHA CI

ggml-ci

* cont : remove -g flag

ggml-ci

b5795

01 Jul 12:13
343b6e9
Compare
Choose a tag to compare
CANN: update aclnnGroupedMatmulV2 to aclnnGroupedMatmulV3 (#14411)

* [CANN]update to aclnnGroupedMatmulV2

Signed-off-by: noemotiovon <[email protected]>

* Support MUL_MAT_ID on 310p

Signed-off-by: noemotiovon <[email protected]>

* fix editorconfig

Signed-off-by: noemotiovon <[email protected]>

---------

Signed-off-by: noemotiovon <[email protected]>

b5794

01 Jul 11:42
6a746cf
Compare
Choose a tag to compare
vulkan: Split large mul_mat_id to fit in shared memory (#14451)

b5793

01 Jul 11:40
eff5e45
Compare
Choose a tag to compare
add GELU_ERF (#14455)

b5792

01 Jul 11:19
Compare
Choose a tag to compare
ggml : remove trailing whitespace (#0)

b5788

01 Jul 07:50
79b33b2
Compare
Choose a tag to compare
opencl : add GEGLU, REGLU, SWIGLU (#14456)

b5787

30 Jun 17:03
0a5a3b5
Compare
Choose a tag to compare
Add Conv2d for CPU (#14388)

* Conv2D: Add CPU version

* Half decent

* Tiled approach for F32

* remove file

* Fix tests

* Support F16 operations

* add assert about size

* Review: further formatting fixes, add assert and use CPU version of fp32->fp16

b5785

30 Jun 14:26
5dd942d
Compare
Choose a tag to compare
metal : disable fast-math for some cpy kernels (#14460)

* metal : disable fast-math for some cpy kernels

ggml-ci

* cont : disable for q4_1

ggml-ci

* cont : disable for iq4_nl

ggml-ci

b5784

30 Jun 13:10
a7417f5
Compare
Choose a tag to compare
ggml-cpu: sycl: Re-enable exp f16 (#14462)

b5783

30 Jun 11:51
eb3fa29
Compare
Choose a tag to compare
test-backend-ops : disable llama test (#14461)