ggml-org/llama.cpp

Pull Requests Commits

llama : use llm_build_granite for minicpm (#13911)

zkh2016 committed 1 year ago

Verified 2c90da4c

cmake: Guard GGML_CPU_ALL_VARIANTS by architecture (#13890)

ckastner committed 1 year ago

Verified ec9e0301

llama : add support for jina-reranker-v2 (#13900)

CISC committed 1 year ago

Verified e83ba3e4

gguf-py : add support for sub_type (in arrays) in GGUFWriter add_key_value method (#13561)

CISC committed 1 year ago

Verified 2b131621

arm64: optimize q4_k_q8_k kernel with i8mm (#13886)

cyb70289 committed 1 year ago

Verified 54a2c7a8

cmake: Factor out CPU architecture detection (#13883)

ckastner committed 1 year ago

Verified 21fcc21a

ggml: aarch64: Implement SVE F32 kernels for Mamba Sequential Scan Algorithm (#13882)

vineelabhinav committed 1 year ago

Verified dd8ba934

tests : remove json.hpp from a test (#13880)

ggerganov committed 1 year ago

Verified 66c92061

convert : workaround for AutoConfig dummy labels (#13881)

CISC committed 1 year ago

Verified 5ca82fc1

llama : add RobertaForSequenceClassification reranker support (#13875)

CISC committed 1 year ago

Verified 6385b843

ggml: aarch64: Implement SVE F32 kernels for vector functions (#13843)

vineelabhinav committed 1 year ago

Verified 1b8fb815

gguf-py : fix SafetensorRemote return on undefined size (< 0) (#13841)

Beinsezii committed 1 year ago

Verified 53ae3064

llama : fix KV shift for qwen2vl (#13870)

ngxson committed 1 year ago

Verified 763d06ed

mtmd : move helpers to dedicated library (⚠️ breaking change) (#13866)

ngxson committed 1 year ago

Verified 10961339

ci: disable LLAMA_CURL for Linux cross-builds (#13871)

mtmcp committed 1 year ago

Verified d98f2a35

llama : add support for BertForSequenceClassification reranker (#13858)

huydt84 committed 1 year ago

Verified e0e3aa23

convert: small addition to support LlamaModel (#13838)

huydt84 committed 1 year ago

Verified aa6dff05

server: fix remove 'image_url'/'input_audio' json-object effectlly for 'llama_params' in multimodal-model-mode (#13853)

flyinskyin2013 committed 1 year ago

Verified c962ae33

convert : fix qwen omni conversion (#13859)

ngxson committed 1 year ago

Verified a3938fb5

tests : change umlaut test (#11600)

n00b001 committed 1 year ago

Verified f7873fc6

CUDA: fix FA tg at long context for CC >= 8.9 (#13852)

JohannesGaessler committed 1 year ago

Verified a6824743

convert : fix tensor naming conflict for llama 4 vision (#13836)

ngxson committed 1 year ago

Verified 26b79b6c

CANN: Add SOC TYPE printing in cmake configuration (#13837)

leo-pony committed 1 year ago

Verified 1e8659e6

opencl: add new ops - `argsort`, `div`, `sub`, `addrows`, `sigmoid`, `group_norm` (#13787)

lhez committed 1 year ago

Verified a3c30846

opencl: mark `mul_mat` `f32f32` as supporting non-contiguous tensors (#13790)

lhez committed 1 year ago

Verified 1701d4c5

vulkan: use timestamp queries for GGML_VULKAN_PERF (#13817)

jeffbolznv committed 1 year ago

Verified bef81763

cmake : add llama-cparams.cpp to build (#13832)

ggerganov committed 1 year ago

Verified 34b7c043

SYCL: add gelu_erf kernel (#13749)

qnixsynapse committed 1 year ago

Verified f3101a8c

ggerganov committed 1 year ago

1c49c70d

ggml : add ggml_repeat_4d (#13824)

ngxson committed 1 year ago

Verified a8ea03d8

Newer Older