ggerganov/llama.cpp

Pull Requests Commits
cuda : optimize argmax (#10441)
slaren committed 1 year ago
Verified a5e47592
llama : handle KV shift for recurrent models (#10402)
ggerganov committed 1 year ago
Verified 1bb30bf2
sync : ggml
ggerganov committed 1 year ago
Verified 87a533be
ggml/sched : do not skip views in pre-assignments
slaren committed 1 year ago
Verified 59b91728
ggml-opt: fix data corruption (ggml/1022)
JohannesGaessler committed 1 year ago
Verified 02e4eaf2
vulkan: predicate max operation in soft_max shaders/soft_max (#10437)
jeffbolznv committed 1 year ago
Verified 9abe9eea
cmake: add link dependencies to cmake find pkg (#10433)
bandoti committed 1 year ago
Verified f95caa79
llama : add .clang-format file (#10415)
slaren committed 1 year ago
Verified fab5d30f
vulkan: copy iq4_nl LUT into shared memory (#10409)
jeffbolznv committed 1 year ago
Verified 8fd4b7fa
vulkan: further optimize mul_mat_vec using larger loads (#10387)
jeffbolznv committed 1 year ago
Verified 1bacb9f6
update rel to 4040 (#10395)
NeoZhangJianyu committed 1 year ago
Verified ad21c9e1
Fix missing file renames in Makefile due to changes in commit ae8de6d50a (#10413)
avdg committed 1 year ago
Verified 3952a221
add cmake rvv support (#10411)
lhpqaq committed 1 year ago
Verified 42ae10bb
sync : ggml
ggerganov committed 1 year ago
9fe0fb06
metal : fox offset integer overflows in im2col (ggml/1015)
pminev committed 1 year ago
611fabd7
metal : add `GGML_UNARY_OP_ELU` kernel (ggml/1018)
PABannier committed 1 year ago
12b0ad95
cmake: force MSVC compiler charset to utf-8 (#9989)
MakeDecisionWorth committed 1 year ago
Verified 342397dc
Add required ggml-base and backend libs to cmake pkg (#10407)
bandoti committed 1 year ago
Verified 2a11b6b0
cuda : fix CUDA_FLAGS not being applied (#10403)
slaren committed 1 year ago
Verified 3ee6382d
llama : add check for KV cache shifts (#10401)
ggerganov committed 1 year ago
Verified 8e752a77
llama : add OLMo November 2024 support (#10394)
2015aroras committed 1 year ago
Verified a88ad007
sycl : Add option to set the SYCL architecture for all targets (#10266)
Rbiessy committed 1 year ago
Verified 2a1507c1
vulkan: Optimize soft_max (#10301)
jeffbolznv committed 1 year ago
Verified b3e58598
sycl: Revert MUL_MAT_OP support changes (#10385)
Alberto Cabrera Pérez committed 1 year ago
Verified 557924f2
cuda : only use native when supported by cmake (#10389)
slaren committed 1 year ago
Verified d3481e63
Skip searching root path for cross-compile builds (#10383)
bandoti committed 1 year ago
Verified 531cb1c2
vulkan: remove use of null initializer (#10372)
jeffbolznv committed 1 year ago
Verified f139d2ea
flake.lock: Update (#10346)
ggerganov committed 1 year ago
Verified 2eb76b2a
Vulkan: Fix device info output format specifiers (#10366)
0cc4m committed 1 year ago
Verified 9b75f03c
docker: use GGML_NATIVE=OFF (#10368)
JohannesGaessler committed 1 year ago
Verified 75207b3a
Terms Of Use Privacy Policy Refund Policy Impressum