ggml-org/llama.cpp

Pull Requests Commits

ggerganov committed 11 days ago

Verified 6ecba0d0

ggerganov committed 11 days ago

Verified 94bfa780

ggerganov committed 11 days ago

Verified 3e0a3e86

ggerganov committed 11 days ago

Verified bd48a0ac

ggerganov committed 11 days ago

Verified ab6f1122

ggerganov committed 11 days ago

Verified faad7d47

arg : add shorthand for --backend-sampling

ggerganov committed 11 days ago

Verified 23e8bb40

Merge remote-tracking branch 'upstream/master' into backend-sampling

danbev committed 11 days ago

Verified ebfe545c

webui: fix prompt progress ETA calculation (#18468)

ngxson committed 12 days ago

Verified 51a48720

Webui/prompt processing progress (#18300)

ServeurpersoCom committed 12 days ago

Verified c9a3b40d

CUDA: fix replacment of bad archs in CMake (#18457)

JohannesGaessler committed 12 days ago

Verified 0bd1212a

server : Cmdline arg -to changes http read timeout from current 600sec default (#18279)

wbtek committed 12 days ago

Verified 5b1248c9

contributing: tighten AI usage policy (#18388)

ngxson committed 12 days ago

Verified 3595ae59

android: routine maintenance - Dec 2025 (#18338)

naco-siren committed 12 days ago

Verified c1366056

server : handle closed connection for tasks (#18459)

ggerganov committed 12 days ago

Verified 2a85f720

model-conversion : add device option to embd run orig model (#18386)

danbev committed 12 days ago

Verified 7cbec34a

retrieval : use at most n_seq_max chunks (#18400)

hectorem2 committed 12 days ago

Verified 0c898640

common: fix return value check for setpriority (#18412)

o7si committed 12 days ago

Verified daa242df

CUDA: Blackwell features for non-native builds (#18436)

JohannesGaessler committed 12 days ago

Verified e70e640d

cuda: fix race condition in cumsum (#18448)

am17an committed 12 days ago

Verified 5fa66c6e

ci : re-enable rocm build on amd64 (#18439)

TimNN committed 12 days ago

Verified 382808c1

HIP: Use mmq on MFMA devices for MUL_MAT_ID in cases where a lot of splits would be generated (#18202)

IMbackK committed 13 days ago

Verified 4ffc47cb

model : Plamo3 support (#17304)

mmnga committed 13 days ago

Verified 9c675c71

Revert "ggml-cuda: use CMAKE_CUDA_ARCHITECTURES if set when GGML_NATIVE=ON (#18413)" (#18426)

am17an committed 13 days ago

Verified 07a0c4ba

rpc: fix segfault on invalid endpoint format (#18387)

o7si committed 13 days ago

Verified 60f17f56

llama-fit-params: fix step size for last device (#18415)

JohannesGaessler committed 13 days ago

Verified f8d561eb

github: update issue templates [no ci] (#18410)

JohannesGaessler committed 13 days ago

Verified e59efe6a

mtmd: clarify that we no longer accept AI-generated PRs (#18406)

ngxson committed 13 days ago

Verified cffa5c46

cmake: Added more x86_64 CPU backends when building with `GGML_CPU_ALL_VARIANTS=On` (#18186)

bberberov committed 13 days ago

Verified 94de74e7

ggml : include cub/cub.cuh instead of block_scan.cuh

danbev committed 13 days ago

Verified 060c0a58

Older