ggerganov/llama.cpp

Pull Requests Commits

fix compile issues

0cc4m committed 27 days ago

562b5573

0cc4m committed 27 days ago

22be6773

0cc4m committed 27 days ago

b72b4e59

add dmabuf capability check

0cc4m committed 27 days ago

76c44178

add reverse order tests for dmabuf

0cc4m committed 27 days ago

7d3f7033

skip dmabuf_p2p when one device is nvidia, due to driver crashes

0cc4m committed 27 days ago

d49112ce

improve output device consistency

0cc4m committed 27 days ago

5e9b4141

catch driver issues in benchmarks

0cc4m committed 27 days ago

4b35cf88

add host dmabuf p2p test

0cc4m committed 27 days ago

f08f1cad

output device group info

0cc4m committed 27 days ago

b39d8bfa

add device group test

0cc4m committed 27 days ago

1e9c9bd1

clean up tests, add dma_buf test

0cc4m committed 27 days ago

e1b0bd7a

0cc4m committed 27 days ago

efdc1cae

ggerganov committed 27 days ago

81191aff

ggml : bump version to 0.22.0 (ggml/1607)

ggerganov committed 27 days ago

93882361

grammar : parse \- in char classes as literal hyphen (#27591)

NIXKnight committed 27 days ago

Verified eb25b726

sycl : mark tq2_0 as not supported (#27660)

arthw committed 27 days ago

Verified 814d84bc

webgpu : fix handling of infinity values during ARGSORT and TOP_K (#27538)

fairydreaming committed 27 days ago

Verified 5ea87dda

metal : per-device tuned (Q, NE) for flash-attn vec (#26570)

forforever73 committed 28 days ago

Verified f280b269

metal: per-op source split + parallel compile (#26561)

forforever73 committed 28 days ago

Verified b615f5b4

misc : read repetition_penalty from generation_config.json (#27659)

tdakhran committed 28 days ago

Verified b3c3b96a

tests : disable DOTS3NOTE arch test for WebGPU (#27654)

fairydreaming committed 28 days ago

Verified 75844307

convert: fix GLM regression in index_tensors (#27655)

jacekpoplawski committed 28 days ago

Verified 71cc86fa

ggml : shorten virtual device naming in CUDA and Metal (#27608)

ggerganov committed 28 days ago

Verified a14dba68

webgpu : reorder includes since V that appears in common_decls.tmpl may be defined as K in flash_attn_decls.tmpl if KV_OVERLAP (#27545)

fairydreaming committed 28 days ago

Verified c1c766da

mtmd: video: fix moov atom at the end of file (#27596)

ngxson committed 28 days ago

Verified 160c6b0b

ci : apply ccache-clear with older/min/dry-run to all ccache jobs (#27602)

ggerganov committed 28 days ago

Verified 985b1491

ggml : fix ggml_clamp (#27644)

ggerganov committed 28 days ago

Verified 6036c635

mamba2 : Flatten in/out projections to dispatch GEMM instead of GEMV (#27513)

pskrunner14 committed 28 days ago

Verified a130532a

Deepseek 4: `-sm tensor` (#26490)

am17an committed 28 days ago

Verified bf0a29cc

Older