ngxson/llama.cpp

Pull Requests Commits

add ref to the PR

ngxson committed 243 days ago

3db4cb0e

llama : fix KV shift for qwen2vl

ngxson committed 243 days ago

9822f2ce

convert: small addition to support LlamaModel (#13838)

huydt84 committed 243 days ago

Verified aa6dff05

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

flyinskyin2013 committed 243 days ago

Verified c962ae33

convert : fix qwen omni conversion (#13859)

ngxson committed 243 days ago

Verified a3938fb5

tests : change umlaut test (#11600)

n00b001 committed 243 days ago

Verified f7873fc6

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

JohannesGaessler committed 243 days ago

Verified a6824743

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

ngxson committed 243 days ago

Verified 26b79b6c

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

leo-pony committed 243 days ago

Verified 1e8659e6

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

lhez committed 244 days ago

Verified a3c30846

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

lhez committed 244 days ago

Verified 1701d4c5

vulkan: use timestamp queries for GGML_VULKAN_PERF (#13817)

jeffbolznv committed 244 days ago

Verified bef81763

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

ggerganov committed 244 days ago

Verified 34b7c043

SYCL: add gelu_erf kernel (#13749)

qnixsynapse committed 244 days ago

Verified f3101a8c

ggerganov committed 244 days ago

1c49c70d

ggml : add ggml_repeat_4d (#13824)

ngxson committed 244 days ago

Verified a8ea03d8

ggml : riscv: add xtheadvector support (#13720)

xctan committed 244 days ago

Verified 05f6ac62

mtmd : support Qwen 2.5 Omni (input audio+vision, no audio output) (#13784)

ngxson committed 244 days ago

Verified bc583e3c

docs: remove link for llama-cli function calling (#13810)

mtmcp committed 244 days ago

Verified 72b090da

ggml-cpu: x86 feature detection is specific to x86 (#13811)

ckastner committed 244 days ago

Verified 7fe03e74

ggml : allow CUDA graphs when using pipeline parallelism (#13814)

slaren committed 244 days ago

Verified 952f3953

kv-cells : track min/max used cells and per-sequence positions (#13808)

ggerganov committed 244 days ago

Verified 81713121

sampling : make sure samplers return at least 1 token (#13822)

ggerganov committed 244 days ago

Verified f9cd6839

llama : validate seq id batch input (#13809)

ggerganov committed 244 days ago

Verified 4f81b33e

server: --offline mode (#13804)

ochafik committed 245 days ago

Verified cdf94a18

scripts : add option to compare commits in Debug (#13806)

ggerganov committed 245 days ago

Verified a26c4cc1

cuda : avoid cuGetErrorString (#13791)

ggerganov committed 245 days ago

Verified 4265a87b

SYCL: Add non contiguous support in RMS_NORM and NORM kernels (#13611)

qnixsynapse committed 245 days ago

Verified 6f180b91

server: fix streaming crashes (#13786)

ochafik committed 245 days ago

Verified 03f582ae

examples/training: Fix file name in README (#13803)

standby24x7 committed 245 days ago

Verified 88c125f2

Older