ngxson/llama.cpp

Pull Requests Commits

rm redundant data member

ngxson committed 268 days ago

dbb257c1

fix prompt_modified

ngxson committed 269 days ago

cd5dc6b8

move hash to user-define ID (fixed)

ngxson committed 270 days ago

d3c3e20c

shared_ptr for mtmd_image_tokens

ngxson committed 270 days ago

58c47674

mtmd : ability to calc image hash

ngxson committed 271 days ago

7ac0b7b7

mtmd : add more api around mtmd_image_tokens

ngxson committed 271 days ago

a46b6db6

`tool-call`: fix non-tool-calling grammar crashes w/ Qwen / Hermes 2 templates (#12900)

ochafik committed 271 days ago

Verified b6930ebc

common : Define cache directory on FreeBSD (#12892)

yurivict committed 271 days ago

Verified 68b08f36

sycl: Support sycl_ext_oneapi_limited_graph (#12873)

EwanC committed 271 days ago

Verified 578754b3

contrib: support modelscope community (#12664)

tastelikefeet committed 271 days ago

Verified b2034c2b

llama-model : add Glm4Model implementation for GLM-4-0414 (#12867)

zRzRzRzRzRzRzR committed 271 days ago

Verified 06bb53ad

clip : use smart pointer (⚠️ breaking change) (#12869)

ngxson committed 271 days ago

Verified 0c509239

SYCL: Add fp16 type support to unary op kernels (#12788)

qnixsynapse committed 271 days ago

Verified fccf9cae

convert : Llama4 RoPE fix (#12889)

danielhanchen committed 271 days ago

Verified ec6c09d0

ci : Replace freediskspace to free_disk_space in docker.yml (#12861)

yeahdongcn committed 271 days ago

Verified 8ac9f5d7

xcf : add check for visionos build version (#12854)

danbev committed 271 days ago

Verified 12e9158f

convert : proper tensor name mapping for llama4 (#12870)

ngxson committed 271 days ago

Verified 5b1f13cb

llama : correct rms norm for llama 4 (#12882)

ngxson committed 271 days ago

Verified 8b91d535

ggml: fix compilation error s390x (#12848)

taronaeo committed 271 days ago

Verified 0fed24c3

ggerganov committed 272 days ago

47ba87d0

tests : fix init order (#0)

ggerganov committed 272 days ago

1d2b6134

ggerganov committed 272 days ago

eb420e11

ggml: don't include arm_neon.h when using CUDA 12 with ARM Neon (ggml/1187)

cmdr2 committed 272 days ago

cb79c2e7

ggml : add bilinear upscale support (ggml/1185)

slaren committed 272 days ago

fe92821e

ggml : add more generic custom op, remove deprecated custom ops (ggml/1183)

slaren committed 272 days ago

459895c3

scripts : fix sync-ggml-am.sh

ggerganov committed 272 days ago

e4bf72d6

llava : introduce libmtmd (#12849)

ngxson committed 272 days ago

Verified 8b9cc7cd

convert : ability to lazy-load safetensors remotely without downloading to disk (#12820)

ngxson committed 272 days ago

Verified 64eda5de

CANN: Support more ops (#12841)

noemotiovon committed 272 days ago

Verified fe5b78c8

Fixes #12823 (#12830)

mehendarkarprajwal committed 273 days ago

Verified 11d07e1e

Older