ggml-org/llama.cpp

Pull Requests Commits

metal : remove stray return

ggerganov committed 68 days ago

Verified d441c31b

ngxson committed 68 days ago

cdb3deae

improve debugging message

ngxson committed 68 days ago

b67a168f

cont : remove obsolete comment [no ci]

ggerganov committed 68 days ago

Verified 29330dcb

clip : print more detailed op support info during warmup

ggerganov committed 68 days ago

Verified bdb43f6e

implement "auto" mode for clip flash attn

ngxson committed 68 days ago

b4955f0a

Merge branch 'master' into gg/clip-fa

ngxson committed 68 days ago

19116a4b

webui: auto-refresh /props on inference start to resync model metadata (#16784)

ServeurpersoCom committed 68 days ago

Verified 2f68ce7c

webui: add HTML/JS preview support to MarkdownContent with sandboxed iframe (#16757)

ServeurpersoCom committed 69 days ago

Verified e4a71599

vendor : update cpp-httplib to 0.27.0 (#16846)

angt committed 69 days ago

Verified dd5e8cab

mtmd: refactor preprocessing + support max/min pixels (#16878)

ngxson committed 69 days ago

Verified cf659bbb

Add a setting to display message generation statistics (#16901)

allozaur committed 69 days ago

Verified d8b860a2

webui: recognize AsciiDoc files as valid text files (#16850)

jhradilek committed 69 days ago

Verified 1ae74882

common : allow --system-prompt-file for diffusion-cli (#16903)

CISC committed 69 days ago

Verified 961660b8

codeowners : update after refactor (#16905)

CISC committed 69 days ago

Verified 74fef412

vulkan: Fix multi_add invalid descriptor usage (#16899)

jeffbolznv committed 69 days ago

Verified 5d8bb900

vulkan: fuse mul_mat+add and mul_mat_id+add_id (#16868)

jeffbolznv committed 69 days ago

Verified 2e76e013

CUDA: Remove unneded bias/gate dims in fused mmvq (#16858)

ORippler committed 69 days ago

Verified d3dc9dd8

refactor : llama-model.cpp (#16252)

pwilkin committed 69 days ago

Verified bea04522

model : Minimax M2 (#16831)

pwilkin committed 69 days ago

Verified 0de0a015

model : add Granite Hybrid nano types (#16896)

giuseppe committed 69 days ago

Verified e58d5856

CUDA: Volta tensor core support for MMF (#16843)

JohannesGaessler committed 70 days ago

Verified 31c511a9

ggerganov committed 70 days ago

6d39015a

CUDA: add expert reduce kernel (#16857)

am17an committed 70 days ago

Verified 4146d6a1

batch : fix consistency checks for the input positions (#16890)

ggerganov committed 70 days ago

Verified 8da3c0e2

server : don't print user inputs to console (#16871)

ggerganov committed 70 days ago

Verified c22473b5

server : fix typos in server.cpp comments [no ci] (#16883)

danbev committed 70 days ago

Verified 0f715b4e

vulkan: disable spirv-opt for rope shaders (#16872)

jeffbolznv committed 70 days ago

Verified d2d931f1

vulkan: Fix crash when FP16 mul_mat accumulation is not supported (#16796)

rillomas committed 70 days ago

Verified 2976b037

vulkan: fix shmem overrun in mmq id shader (#16873)

0cc4m committed 70 days ago

Verified d2a2673d

Older