ggerganov/llama.cpp

Pull Requests Commits

server: add oai-compat /v1/audio/speech endpoint

ngxson committed 1 hour ago

384cbf42

mtmd: enable cmake unity build (#31760)

ngxson committed 2 hours ago

Verified e14a7797

tests : drop u8path() (#31755)

CISC committed 3 hours ago

Verified d38bea34

ci: fix build-cmake-pkg ccache (#31705)

netrunnereve committed 4 hours ago

Verified 186eab1c

llama: fix kv shifting, add tests, add per-model build_rope (#31683)

ngxson committed 5 hours ago

Verified 6023692b

llama, mtmd: (chatterbox) add use_classifier_free_guidance as gguf metadata (#31698)

ngxson committed 6 hours ago

Verified b1c902f0

models : recursive fix (#31749)

CISC committed 6 hours ago

Verified e8df20d2

model : add Ripple (ricursiveai) (#31692)

notforu777 committed 7 hours ago

Verified fecdc81e

model: Add support for the xing4_0 (#31589)

shuxiaoqiong committed 9 hours ago

Verified 45b6604b

ui: prevent a hidden sidebar from intercepting clicks (#31712)

ntowle committed 11 hours ago

Verified e30ad80f

llama: refactoring: move model to class member of llm_graph_context (#31733)

ngxson committed 11 hours ago

Verified e0f96a1a

server : simplify windows MCP command resolution (#31728)

angt committed 11 hours ago

Verified 468fde90

vendor : update cpp-httplib to 0.65.0 (#31731)

angt committed 11 hours ago

Verified 1416f76a

ggml-cuda: sanitize post-bias NaNs in MoE selection (#31005)

ArberSephirotheca committed 13 hours ago

Verified 10d4846a

vulkan: Implement test-fusion interface (#30769)

jeffbolznv committed 13 hours ago

Verified 61d44723

CUDA: use highest compiled arch in get_mmvq_mmid_max_batch (#31679)

ManuelPDM committed 13 hours ago

Verified 016c1d91

vulkan: RDNA3/4 Large-Tile Split-K Optimization (#30280)

Alex-JP-93 committed 13 hours ago

Verified da3416d8

gguf-py : external chat templates have precedence over tokenizer_config (#31717)

CISC committed 13 hours ago

Verified cc78da97

metal : fuse DeepSeek-V4/V4.1 and GLM MoE routing (#31615)

ggerganov committed 14 hours ago

Verified ea40a3a1

CUDA/HIP: matrix-core (MFMA) lightning indexer for CDNA2 (gfx90a) (#29050)

jadenmach2 committed 15 hours ago

Verified 15e50990

CUDA: linearize MoE data layout for MMQ/MMF (#31047)

JohannesGaessler committed 15 hours ago

Verified 69ee2ecc

vulkan: FWHT kernels for block widths above 512 (#29772)

bri-prism committed 15 hours ago

Verified 2bc46baa

vulkan : check combined K+V f16 scratch size in FA dequant path (#30487)

nisbenz committed 15 hours ago

Verified 3b83d12b

llama-server: Use reference-counted checkpoints instead of deep-copies (#30137)

cwriter committed 15 hours ago

Verified e152332a

ggml : support `truncate=false` for YaRN (#30206)

tokinasin committed 15 hours ago

Verified 1fa7f1c4

conversion : resolve quantized_layers from source tensor names (#31611)

danbev committed 16 hours ago

Verified 9c9074e1

metal : use mul_mm for 32 <= K < 64 (#30982)

vieduy committed 16 hours ago

Verified f7b41cca

server : reject non-object session updates (#31714)

angt committed 17 hours ago

Verified c55b0948

fit : account for skipped NextN layers when setting n_gpu_layers (#31569)

am17an committed 17 hours ago

Verified 2c192444

webgpu : add MEAN support (#31591)

LeON-Nie-code committed 17 hours ago

Verified 553d2c2e

Older