ggml-org/llama.cpp

Pull Requests Commits

server: explicit opt-in tools via non-loopback addr, check host port

ngxson committed 16 minutes ago

1b365f05

ui: move js sandbox setting out of "developer" (#32054)

ngxson committed 1 hour ago

Verified 7bcbc04f

docs: update image link in README.md (#32038)

inimaz committed 1 hour ago

Verified dd00306b

ui: keep downloads in Discover during search and fix their progress with a draft (#31467)

allozaur committed 1 hour ago

Verified 1d819927

ui: take the logo and favicons from the llama-brand package (#32039)

allozaur committed 1 hour ago

Verified bd717e08

server: fix --cors-origins handling + protect against DNS rebinding (#32098)

ngxson committed 3 hours ago

Verified 2cbe0ec4

ci: move test models to ggml-org-dev (#32096)

ngxson committed 3 hours ago

Verified 664f031f

metal : prevent out-of-bounds reads in SOLVE_TRI kernel (#32103)

realMoana committed 4 hours ago

Verified 4180722e

vulkan: support batched tensor copy API (#31418)

0cc4m committed 4 hours ago

Verified 055c3c26

vendor : update LibreSSL to 4.4.1 [no ci] (#32106)

cabelo committed 4 hours ago

Verified 2656b71a

llama : reuse graph result instances across scheduler reserves (#32089)

ggerganov committed 6 hours ago

Verified cb44fdb1

jinja : support chained comparisons and argument unpacking (#32072)

aldehir committed 6 hours ago

Verified cd2db800

chat: allow TTS model to provide custom text normalizer via jinja (#32053)

ngxson committed 7 hours ago

Verified f7b886fe

ci : remove dead snakes (#32084)

CISC committed 9 hours ago

Verified 956591f4

llama : remove non-fused GDN and ubatch-dependent graph nodes (#32078)

ggerganov committed 10 hours ago

Verified ccd360f0

devops : add missing MKL runtime DLLs to the Windows SYCL release package (#32005)

Asahi-Prv committed 10 hours ago

Verified b326dd67

vendor: update cpp-httplib to 0.66.0 (#32049)

cabelo committed 11 hours ago

Verified b8abe41a

backend : re-instantiate CUDA graphs when the cluster dims of a kernel change (#31284)

tdakhran committed 12 hours ago

Verified 2e325dcd

test-fp8.cpp: use only the public ggml API (#32083)

ORippler committed 12 hours ago

Verified 1648d641

ROCm/HIP: fixed BLAS workspace instead of rocBLAS allocating one per call (#31838)

fox5611 committed 12 hours ago

Verified 43243363

quantize: fix imatrix GGUF filepath metadata leak (#32000)

JamieBordoni committed 12 hours ago

Verified 71a6a114

cpu: accumulate flash attention V in F32 (#29810)

ServeurpersoCom committed 13 hours ago

Verified 7f24aacd

webgpu: add PAD_REFLECT_1D support (#31877)

LeON-Nie-code committed 13 hours ago

Verified af6b5243

vulkan: allow dequantized KV optimization with CM2 (#31526)

jeffbolznv committed 13 hours ago

Verified d2b3eb9b

vulkan: allocate the device fault vendor binary buffer (#32008)

figkcoci-debug committed 14 hours ago

Verified eeb21047

tests : use only the public ggml API (#32077)

ORippler committed 14 hours ago

Verified 03d6c824

ui: open the virtual keyboard in the chat editor on Android (#32060)

ServeurpersoCom committed 14 hours ago

Verified adde94c1

cuda : use MMVF for BF16 matrices of few rows at small batch sizes (#31959)

am17an committed 15 hours ago

Verified 459ab1cd

cuda : add sm70 FlashAttention config case for DKQ=256, DV=256, ncols=64 (#27997)

mistrjirka committed 15 hours ago

Verified 77e2db8c

cuda : fattn mma head sizes 40 72 (#30285)

sihanyu03 committed 15 hours ago

Verified 9d7b2d73

Older