Skip to content

Releases: allozaur/llama.cpp

b6344

01 Sep 14:45
02c1813

Choose a tag to compare

Vulkan: Add Integer Dot Product mul_mat_vec shader for legacy quants …

b6341

01 Sep 07:30
b66df9d

Choose a tag to compare

CUDA: fix build error from ambiguous __half conversions in conv2d (#1…

b6259

24 Aug 00:23
710dfc4

Choose a tag to compare

CUDA: fix half2 -> half conversion for HIP (#15529)

b6209

19 Aug 23:22
fb22dd0

Choose a tag to compare

opencl: mark `argsort` unsupported if cols exceed workgroup limit (#1…

b6160

14 Aug 13:03
d32e03f

Choose a tag to compare

server : add SWA checkpoints (#15293)

* server : add SWA checkpoints

ggml-ci

* cont : server clean-up

* server : handle state restore fails

* llama : add extended llama_state_seq_ API

* server : do not make checkpoints if --swa-full

ggml-ci

* llama : remove flags value for NONE

* server : configure number of SWA checkpoints with CLI arg

ggml-ci

* args : fix scope of new argument

b6075

02 Aug 22:57
5c0eb5e

Choose a tag to compare

opencl: fix adreno compiler detection logic (#15029)

b6059

01 Aug 19:17
c76b420

Choose a tag to compare

vendor : update vendored copy of google/minja (#15011)

* vendor : update vendored copy of google/minja

Signed-off-by: Lennart Austenfeld <[email protected]>

* Re-remove trailing whitespace

Signed-off-by: Lennart Austenfeld <[email protected]>

* Remove another trailing whitespace

Signed-off-by: Lennart Austenfeld <[email protected]>

---------

Signed-off-by: Lennart Austenfeld <[email protected]>

b6027

29 Jul 19:24
aa79524

Choose a tag to compare

HIP: remove the use of __HIP_PLATFORM_AMD__, explicitly support only …

b6026

29 Jul 18:02
b77d111

Choose a tag to compare

HIP: add GGML_HIP_MMQ_MFMA option to allow disableing the MFMA path. …