ngxson/llama.cpp

Pull Requests Commits

fix crash on windows

ngxson committed 4 hours ago

95b79963

fix unset_reserved_args

ngxson committed 4 hours ago

11f8109b

skip loading custom_models if not specified

ngxson committed 9 hours ago

ac6f8ca1

Merge branch 'master' into xsn/refactor_server_preset

ngxson committed 11 hours ago

7cbb2d2c

ggml-cpu: extend support for RVV floating-point kernels (#17318)

taimur-10x committed 11 hours ago

Verified f716588e

arg: fix ASAN error on sampler_type_names empty (#18167)

ngxson committed 12 hours ago

Verified 4d1316c4

also filter out server-controlled args before to_ini()

ngxson committed 12 hours ago

4004c477

fix empty mmproj

ngxson committed 12 hours ago

60ec94ed

fix neg arg handling

ngxson committed 12 hours ago

0d04bba5

ngxson committed 13 hours ago

5abab16d

presets: refactor, allow cascade presets from different sources

ngxson committed 13 hours ago

4e475a8b

gguf-py : use copy-on-write mode for localtensor (#18162)

CISC committed 13 hours ago

Verified ec7b9329

remove i_major_dual (#18157)

zhang-hui-yulo committed 14 hours ago

Verified 54189c0d

webui: Fix selecting generated output issues during active streaming (#18091)

allozaur committed 15 hours ago

Verified 9ce64aed

webui: fix chat screen shadow width (#18010)

polydecay committed 15 hours ago

Verified 900316da

llama: offload output layer to GPU first (#18148)

JohannesGaessler committed 18 hours ago

Verified 57c1e056

convert : sort and use file parts from model index if present (#18043)

CISC committed 19 hours ago

Verified 9cff4cc5

llama : Async DirectIO model loading on Linux (#18012)

JTischbein committed 19 hours ago

Verified 4d4f4cac

ggml-hexagon: swiglu_oai operation (#18114)

joeldushouyu committed 1 day ago

Verified 0a0bba05

convert : force patch_merger tensors to f16/f32 (#18124)

CISC committed 1 day ago

Verified 5166aaf8

server: (webui) add --webui-config (#18028)

ServeurpersoCom committed 1 day ago

Verified 6ce3d857

server: (router) disable SSL on child process (#18141)

ngxson committed 1 day ago

Verified e85e9d76

llama-fit-params: fix memory print (#18136)

JohannesGaessler committed 1 day ago

Verified 8dcc3662

webui: fix chat header width when sidebar is closed (#17981)

polydecay committed 1 day ago

Verified d37fc935

ggml-hexagon: gelu operation (#17921)

joeldushouyu committed 1 day ago

Verified 4470a076

common : restore grammar-based rejection sampling (#18137)

ggerganov committed 1 day ago

Verified 4301e273

common: clarify instructions for bug reports (#18134)

JohannesGaessler committed 1 day ago

Verified a2c199e4

model: fix GLM-ASR-Nano-2512 load error (#18130) (#18142)

HonestQiao committed 1 day ago

Verified 15dd67d8

server: (router) allow child process to report status via stdout (#18110)

ngxson committed 1 day ago

Verified bde461de

Extend run-org-model.py, add (a) batching (b) loading prompt from file (c) multimodal capacity (#18034)

pwilkin committed 1 day ago

Verified 8faa87db

Older