ngxson/llama.cpp

Pull Requests Commits

ngxson committed 47 days ago

5600361c

Apply suggestions from code review

ngxson committed 47 days ago

Verified b185b7fc

ngxson committed 47 days ago

34234a50

improve arch handling

ngxson committed 47 days ago

bf08fcc4

Add Ministral3 support to Mistral format

juliendenize committed 47 days ago

a4f540bc

Merge remote-tracking branch 'mistral/xsn/ministral3' into xsn/ministral3

ngxson committed 47 days ago

55a196f9

llama-graph: avoid expand_forward for fusion (#17633)

am17an committed 48 days ago

Verified 6eea6669

contributing: update guidelines for AI-generated code (#17625)

ngxson committed 48 days ago

Verified ff90508d

cmake : add option to build and link LibreSSL (#17552)

angt committed 48 days ago

Verified 0a4aeb92

model: LFM2-VL fixes (#17577)

tdakhran committed 48 days ago

Verified 2ba71951

clip: fix nb calculation for qwen3-vl (#17594)

ngxson committed 48 days ago

Verified 7f8ef50c

cli: add migration warning (#17620)

ngxson committed 48 days ago

Verified 3c136b21

common : throttle download progress output to reduce IO flush (#17427)

angt committed 48 days ago

Verified beb1f0c5

common: add LLAMA_LOG_FILE env var (#17609)

taronaeo committed 48 days ago

Verified def5404f

ggml: fix: macOS build with `-DGGML_BACKEND_DL=ON` (#17581)

giladgd committed 49 days ago

Verified fa046595

common: update env var name (#17588)

ddh0 committed 49 days ago

Verified 5a6241fe

CUDA: add stream-based concurrency (#16991)

am17an committed 49 days ago

Verified c7af376c

cuda : add error checking for cudaMemcpyAsync in argsort (#17599)

Mahekk357 committed 49 days ago

Verified 00425e2e

vulkan : fix FA mask load with bounds check (coopmat2) (#17606)

Acly committed 49 days ago

Verified 385c3da5

server: move server-context to its own cpp|h (#17595)

ngxson committed 49 days ago

Verified ab49f094

server: explicitly set the function name in lambda (#17538)

haiyuewa committed 49 days ago

Verified 8c32d9d9

common : fix json schema with '\' in literals (#17307)

i-v-s committed 49 days ago

Verified 0874693b

sycl : support to malloc memory on device more than 4GB, update the doc and script (#17566)

arthw committed 49 days ago

Verified 7d2add51

ggml: replace hwcap with riscv_hwprobe for RVV detection (#17567)

ixgbe committed 49 days ago

Verified f698a79c

Vulkan: MMVQ Integer Dot K-Quant and MUL_MAT_ID support (#16900)

0cc4m committed 50 days ago

Verified 47a268ea

vulkan: improve topk perf for large k, fix overflow in unit tests (#17582)

jeffbolznv committed 50 days ago

Verified 59d8d4e9

gguf-py : fix passing non-native endian tensors (editor-gui and new-metadata) (#17553)

AlekseiNikiforovIBM committed 50 days ago

Verified d82b7a7c

common : move all common_chat_parse_* to chat-parser.cpp. (#17481)

dranger003 committed 50 days ago

Verified 03914c7e

server: fix: /metrics endpoint returning JSON-escaped Prometheus format (#17386)

o7si committed 50 days ago

Verified 3ce7a65c

ggml : add GGML_SCHED_NO_REALLOC option to disable reallocations in ggml_backend_sched (#17276)

slaren committed 50 days ago

Verified e072b205

Older