ggerganov/llama.cpp

Pull Requests Commits

common : skip model validation when --help is requested (#17755)

danbev committed 26 days ago

Verified bd4ef134

ggml-cpu : remove asserts always evaluating to false (#17728)

Alcpz committed 26 days ago

Verified 87a2084c

convert: use existing local chat_template if mistral-format model has one. (#17749)

SmartestWashingMachine committed 27 days ago

Verified 3659aa28

cmake : simplify build info detection using standard variables (#17423)

angt committed 27 days ago

Verified 2a73f81f

ci : disable ggml-ci-x64-amd-* (#17753)

CISC committed 27 days ago

Verified 7dba049b

common: use native MultiByteToWideChar (#17738)

angt committed 27 days ago

Verified 83c11715

metal : use params per pipeline instance (#17739)

ggerganov committed 27 days ago

Verified 0d132485

llama : fix sanity checks during quantization (#17721)

ggerganov committed 27 days ago

Verified a67ef0f4

build : move _WIN32_WINNT definition to headers (#17736)

angt committed 27 days ago

Verified ef75a89f

build: enable parallel builds in msbuild using MTT (#17708)

jeffbolznv committed 27 days ago

Verified d8b5cdc4

ggml-cpu: remove duplicate conditional check 'iid' (#17650)

GermanAizek committed 27 days ago

Verified dea9ba27

Add a couple of file types to the text section (#17670)

pwilkin committed 27 days ago

Verified c6d1a00a

convert : support latest mistral-common (fix conversion with --mistral-format) (#17712)

SmartestWashingMachine committed 27 days ago

Verified 424c5794

Use OpenAI-compatible `/v1/models` endpoint by default (#17689)

allozaur committed 27 days ago

Verified e9f94834

webui: Fix zero pasteLongTextToFileLen to disable conversion being overridden (#17445)

awasisto committed 27 days ago

Verified 41c5e02f

CUDA: generalized (mma) FA, add Volta support (#17505)

JohannesGaessler committed 27 days ago

Verified 2e1c9cd8

chat : reserve memory in compute_diffs and improve naming (#17729)

ggerganov committed 27 days ago

Verified 190c4838

server: add router multi-model tests (#17704) (#17722)

ServeurpersoCom committed 27 days ago

Verified e7c2cf13

server : fix bad fmt, size() is a size_type (#17735)

angt committed 27 days ago

Verified 12574910

cmake: explicitly link against crypt32 on non-MSVC Windows builds (#17727)

angt committed 27 days ago

Verified 083e18b1

metal : fix data race in pipeline library (#17731)

ggerganov committed 28 days ago

Verified 3d94e967

ci : remove the build of openeuler-cann in release (#17724)

xuedinge233 committed 28 days ago

Verified 7feb0a10

common : introduce composable PEG parser combinators for chat parsing (#17136)

aldehir committed 28 days ago

Verified 0a8026e7

server: fix duplicate HTTP headers in multiple models mode (#17698)

ServeurpersoCom committed 28 days ago

Verified 5ceed624

ggml webgpu: add support for emscripten builds (#17184)

reeselevine committed 28 days ago

Verified 7ca5991d

ci : move release details to the top visible by default (#17719)

CISC committed 28 days ago

Verified b3e3060f

ggml, llama : use defaulted constructors/destructors (#17649)

GermanAizek committed 28 days ago

Verified 37adc9c6

build: document how to compile with Vulkan using Debian/Ubuntu packages (#17688)

socram8888 committed 28 days ago

Verified 16cc3c60

server: add --media-path for local media files (#17697)

ngxson committed 28 days ago

Verified 13628d8b

mtmd: fix --no-warmup (#17695)

ngxson committed 28 days ago

Verified a96283ad

Newer Older