ggerganov/llama.cpp

Pull Requests Commits

models : guard the expert FFN size fallback in nemotron-h against a zero divisor (#28779)

YanissAmz committed 39 days ago

Verified e49d2c27

ggml-cpu(s390x): guard VXE-only repack helpers (#28775)

bernardladenthin committed 39 days ago

Verified 69780529

tests : fix typo in test-quant-type-selection for nemotron 3 nano (#28835)

mctylr-gh committed 39 days ago

Verified 243a3082

sycl : Fix get mem error (#28227)

arthw committed 39 days ago

Verified b6b003d2

ci : run editorconfig and code-style checks on ubuntu-slim (#28854)

ggerganov committed 39 days ago

Verified c95f8e47

pi : prefer PI_MODEL_NAME env var for model disclosure (#28853)

ggerganov committed 39 days ago

Verified bc52a12b

tests : reduce FA test sizes (#28842)

ggerganov committed 40 days ago

Verified 4a899373

ci : remove leftover command (#28839)

CISC committed 40 days ago

Verified 37b3a9e0

ci : cap test-backend-ops parallel jobs at 2 and add a 3600s timeout (#28833)

ggerganov committed 40 days ago

Verified 002a12ad

vulkan: workaround NV queuesubmit driver bug (#28830)

jeffbolznv committed 40 days ago

Verified f1e44dcc

opencl: apply the noshuffle row-alignment rule to q4_K, q5_K and q8_0, not just q6_K (#28575)

wanghqc committed 40 days ago

Verified 56b9eb28

chat : improve parsing of complex types in qwen3-coder (#28742)

aldehir committed 40 days ago

Verified 790cf51a

common: add LOG_JSON macro to log structured data (#28586)

ngxson committed 40 days ago

Verified 8e330954

common : implement common_schema internal representation for JSON schemas (#28736)

aldehir committed 40 days ago

Verified acecd560

jinja : support dot property integer literals (#28817)

CISC committed 40 days ago

Verified ae9afff8

cmake: leave the timestamp out of precompiled headers on clang (#28816)

ServeurpersoCom committed 40 days ago

Verified 737e0980

ui : add cache (#28802)

angt committed 40 days ago

Verified 3057bb66

server : allow model downloads at model limit fix issue #26809 (#28530)

chenmiaoming committed 41 days ago

Verified 56381e40

ggml-cuda: hip add specific config table for AMD GCN (#27841)

thelittlefireman committed 41 days ago

Verified c8edceb0

server : add missing headers (#28795)

angt committed 41 days ago

Verified e192abb4

vendor : update cpp-httplib to 0.56.0 (#28787)

cabelo committed 41 days ago

Verified 718f7b41

syscl : Handle (fail gracefully) unsupported tq1_0 quants (#28681)

mctylr-gh committed 41 days ago

Verified 2a3005c2

rpc : fix linking when compiling with BUILD_SHARED_LIBS=OFF (#28492)

EAddario committed 41 days ago

Verified f3a33dff

server: frame the router child state command as a whole line (#28747)

ServeurpersoCom committed 41 days ago

Verified c069aa7f

opencl: fix several bugs where the backend aborts (#27630)

wanghqc committed 41 days ago

Verified 8a56aedd

opencl: add bin kernel `kernel_gemm_noshuffle_q4_k_f32_32b_trans_ila_a8_bin` (#28677)

shaofeiqi committed 41 days ago

Verified 07fc9771

webgpu: align tensor bindings to the type block size (#28382)

ServeurpersoCom committed 41 days ago

Verified 3f5e94d7

hexagon: support for multi-device model split (aka row-split) (#28589)

max-krasnyansky committed 41 days ago

Verified eafe15a5

ggml-webgpu: Update to a recent version of Dawn (#28683)

MendyBerger committed 41 days ago

Verified d3146f2b

server: refactor subproc handling (#28555)

ngxson committed 41 days ago

Verified 82d6bb28

Newer Older