ggerganov/llama.cpp

Pull Requests Commits

0cc4m committed 27 days ago

bfc04053

properly check for cm_m1 features

0cc4m committed 28 days ago

1de478bf

show cm maintenance1 marker in device info

0cc4m committed 28 days ago

51443839

enable by default

0cc4m committed 28 days ago

76535993

0cc4m committed 28 days ago

b25101f9

improve cm1 perf

0cc4m committed 28 days ago

75823219

improve cm2 perf

0cc4m committed 28 days ago

7ff8fe48

fix cm2 validation errors

0cc4m committed 28 days ago

d9cedfea

fix nvidia shmem overuse

0cc4m committed 28 days ago

da8d4883

optimize for Ampere

0cc4m committed 28 days ago

094d5670

add gmem version RDNA4

0cc4m committed 28 days ago

94eb47f0

optimize for RDNA4

0cc4m committed 28 days ago

aa5f6ca9

add coopmat maintenance1 variant

0cc4m committed 28 days ago

a4c938c9

optimize further

0cc4m committed 28 days ago

df078fb5

fix shmem bank conflict bottleneck

0cc4m committed 28 days ago

bde3ccf6

vulkan: add cm2 chunked GATED_DELTA_NET shader

0cc4m committed 28 days ago

f200094f

sycl: add grouped MoE XMX GEMM (#29245)

cwriter committed 28 days ago

Verified 847f447c

convert : support Qwen3.5 embedding models (#27920)

SamMalayek committed 28 days ago

Verified 75118a3a

sycl: remove duplicate block-size defines from op headers (#29507)

Titaniumtown committed 28 days ago

Verified 9b4ed0ca

hexagon: enable alloc_buffer_n (#30126)

max-krasnyansky committed 28 days ago

Verified 9c2e0e49

hexagon: Q6_K weight dequant speedup (#30121)

kurquhar committed 28 days ago

Verified 06cad0b9

model : add LiquidAI/d1-omni-600M decision model (#30114)

tdakhran committed 28 days ago

Verified a657f7e9

hexagon: support tiled Q4_K and Q6_K GET_ROWS (#30115)

kurquhar committed 28 days ago

Verified aa5e0092

hexagon: improved GELU accuracy (#30104)

kurquhar committed 28 days ago

Verified 70815103

chat: fix jinja parser for TranslateGemma (#30096)

nbud committed 29 days ago

Verified bd4eeaa0

chat : name tool and argument parser rules by index (#30088)

Frost-54 committed 29 days ago

Verified 5de73343

model : add LiquidAI/d1-3B decision model (#30110)

tdakhran committed 29 days ago

Verified 88dcc460

cuda: FWHT kernels for block widths above 512 (#29100)

bri-prism committed 29 days ago

Verified b86d2f07

mtmd: add cohere2 vision support (#30062)

Terrencezzj committed 29 days ago

Verified 50a6c5cf

llama : add a GPU cache for MoE experts kept in host memory (#29887)

am17an committed 29 days ago

Verified d6cf9acb

Older