 Johannes Gäßler
|
599b3e0cd4
GitHub: ask for more info in issue templates (#10426)
|
1 anno fa |
 leo-pony
|
c18610b4ee
CANN: Support Ascend310P to accelerate F32 and F16 Model (#10216)
|
1 anno fa |
 Diego Devesa
|
a5e47592b6
cuda : optimize argmax (#10441)
|
1 anno fa |
 Georgi Gerganov
|
1bb30bf28c
llama : handle KV shift for recurrent models (#10402)
|
1 anno fa |
|
Georgi Gerganov
|
87a533be57
sync : ggml
|
1 anno fa |
|
slaren
|
59b9172822
ggml/sched : do not skip views in pre-assignments
|
1 anno fa |
|
Johannes Gäßler
|
02e4eaf22f
ggml-opt: fix data corruption (ggml/1022)
|
1 anno fa |
 Jeff Bolz
|
9abe9eeae9
vulkan: predicate max operation in soft_max shaders/soft_max (#10437)
|
1 anno fa |
 bandoti
|
f95caa7954
cmake: add link dependencies to cmake find pkg (#10433)
|
1 anno fa |
|
Diego Devesa
|
fab5d30ff6
llama : add .clang-format file (#10415)
|
1 anno fa |
|
Jeff Bolz
|
8fd4b7fa29
vulkan: copy iq4_nl LUT into shared memory (#10409)
|
1 anno fa |
|
Jeff Bolz
|
1bacb9f625
vulkan: further optimize mul_mat_vec using larger loads (#10387)
|
1 anno fa |
 Neo Zhang Jianyu
|
ad21c9e1f1
update rel to 4040 (#10395)
|
1 anno fa |
 Anthony Van de Gejuchte
|
3952a221af
Fix missing file renames in Makefile due to changes in commit ae8de6d50a (#10413)
|
1 anno fa |
 haopeng
|
42ae10bbcd
add cmake rvv support (#10411)
|
1 anno fa |
|
Georgi Gerganov
|
9fe0fb0626
sync : ggml
|
1 anno fa |
 Plamen Minev
|
611fabd792
metal : fox offset integer overflows in im2col (ggml/1015)
|
1 anno fa |
 PAB
|
12b0ad953a
metal : add `GGML_UNARY_OP_ELU` kernel (ggml/1018)
|
1 anno fa |
 蕭澧邦
|
342397dc7e
cmake: force MSVC compiler charset to utf-8 (#9989)
|
1 anno fa |
|
bandoti
|
2a11b6b094
Add required ggml-base and backend libs to cmake pkg (#10407)
|
1 anno fa |
|
Diego Devesa
|
3ee6382d48
cuda : fix CUDA_FLAGS not being applied (#10403)
|
1 anno fa |
|
Georgi Gerganov
|
8e752a777b
llama : add check for KV cache shifts (#10401)
|
1 anno fa |
 Shane A
|
a88ad007de
llama : add OLMo November 2024 support (#10394)
|
1 anno fa |
 Romain Biessy
|
2a1507c162
sycl : Add option to set the SYCL architecture for all targets (#10266)
|
1 anno fa |
|
Jeff Bolz
|
b3e585988f
vulkan: Optimize soft_max (#10301)
|
1 anno fa |
 Alberto Cabrera Pérez
|
557924f222
sycl: Revert MUL_MAT_OP support changes (#10385)
|
1 anno fa |
|
Diego Devesa
|
d3481e6316
cuda : only use native when supported by cmake (#10389)
|
1 anno fa |
|
bandoti
|
531cb1c233
Skip searching root path for cross-compile builds (#10383)
|
1 anno fa |
|
Jeff Bolz
|
f139d2ea61
vulkan: remove use of null initializer (#10372)
|
1 anno fa |
|
Georgi Gerganov
|
2eb76b2a5e
flake.lock: Update (#10346)
|
1 anno fa |
