ngxson/llama.cpp

Pull Requests Commits

ggml : add ggml_repeat_4d

ngxson committed 232 days ago

6e7a0620

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

mtmcp committed 232 days ago

Verified 72b090da

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

ckastner committed 232 days ago

Verified 7fe03e74

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

slaren committed 232 days ago

Verified 952f3953

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

ggerganov committed 232 days ago

Verified 81713121

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

ggerganov committed 232 days ago

Verified f9cd6839

llama : validate seq id batch input (#13809)

ggerganov committed 232 days ago

Verified 4f81b33e

server: --offline mode (#13804)

ochafik committed 233 days ago

Verified cdf94a18

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

ggerganov committed 233 days ago

Verified a26c4cc1

cuda : avoid cuGetErrorString (#13791)

ggerganov committed 233 days ago

Verified 4265a87b

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

qnixsynapse committed 233 days ago

Verified 6f180b91

server: fix streaming crashes (#13786)

ochafik committed 233 days ago

Verified 03f582ae

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

standby24x7 committed 233 days ago

Verified 88c125f2

`server`: fix format of streamed tool call deltas (diff name, fix id location) (#13800)

ochafik committed 233 days ago

Verified d74e94c1

server: fix regression on streamed non-chat completion w/ stops (#13785)

ochafik committed 233 days ago

Verified f13847cf

examples : allow extracting embeddings from decoder contexts (#13797)

ggerganov committed 233 days ago

Verified 79c137f7

llama : clarify deprecation message (#13794)

ggerganov committed 233 days ago

Verified 22229314

sycl: Add more debug prints (#13640)

Rbiessy committed 233 days ago

Verified 9012eb9b

vulkan: mark IM2COL as supporting non-contig (#13783)

jeffbolznv committed 234 days ago

Verified fef693dc

CANN: Add the basic supports of Flash Attention kernel (#13627)

shibizhao committed 234 days ago

Verified 2d38b6e4

`server`: add `--reasoning-budget 0` to disable thinking (incl. qwen3 w/ enable_thinking:false) (#13771)

ochafik committed 234 days ago

Verified e121edc4

webui : bump max upload file size to 500MB (#13779)

ngxson committed 234 days ago

Verified 2f099b51

tests : improve UGM tokenizer test coverage (#13773)

CISC committed 234 days ago

Verified aa50ba46

kv-cache : rework kv_cell (#13706)

ggerganov committed 234 days ago

Verified de2ef53a

rpc : Fix build on OpenBSD (#13541)

percypiper committed 234 days ago

Verified c508256d

mtmd : add support for Qwen2-Audio and SeaLLM-Audio (#13760)

ngxson committed 234 days ago

Verified 40aaa8a4

docs : add Moondream2 pre-quantized link (#13745)

ddpasa committed 234 days ago

Verified a08c1d28

server: fix/test add_generation_prompt (#13770)

ochafik committed 234 days ago

Verified d785f9c1

llama : add support for Qwen3 MoE tied word embeddings (#13768)

estibi committed 234 days ago

Verified 4032ca40

SYCL: revert "sycl: simplify bin_bcast_kernel (#13383)" (#13752)

qnixsynapse committed 234 days ago

Verified 515fdbf7

Older