ngxson/llama.cpp

Pull Requests Commits

server: temporary skip model downloading API test

ngxson committed 113 days ago

48badc50

ggml-cpu: use UE4M3 LUT in ARM NVFP4 dot product (#25331)

ragz4125 committed 113 days ago

Verified 20a04b22

ggml-cpu: Enable tiled matmul on AIX (#25199)

shalinib-ibm committed 114 days ago

Verified 3b4fca11

vulkan: fix 32-bit integer overflow in CEIL_DIV (#25245)

hokanosekai committed 114 days ago

Verified 86961efd

ui: restore Ctrl+B sidebar toggle shortcut (#25307)

ServeurpersoCom committed 114 days ago

Verified d80e8785

scripts : use HF_TOKEN when downloading UI assets (#25280)

angt committed 114 days ago

Verified 48719618

ggml-hip: enable -ffast-math for HIP builds (#23862)

a-huk committed 114 days ago

Verified d06ddd35

ui: fake 200 for proxy DELETE req (#25298)

ngxson committed 114 days ago

Verified 898b0885

ggml-cuda: optimize conv_transpose_1d indexing (#25310)

adavyas committed 114 days ago

Verified 72874f55

Fix stale tensor-split params for draft models (#24814)

RedToasty committed 114 days ago

Verified 2da66861

abort if we see a multi buffer (#25276)

netrunnereve committed 114 days ago

Verified 3e5036fb

ggml : fix tensor-parallel + -ncmoe crash on MoE models (#25028)

liminfei-amd committed 114 days ago

Verified 4b2a0cde

ggml: Update VMM Pool allocation ggml-cuda.cu - Turing P2P access fix (fixes #24489) (#24491)

VexxieCode committed 114 days ago

Verified 7a63fded

cuda : concat implementation for quantized types (#25303)

fairydreaming committed 114 days ago

Verified 78d2f524

llama : add guard for K/V rotation input when buffer is unallocated (#25215)

liminfei-amd committed 115 days ago

Verified a4107133

ui: add sync blocks so display/behavior settings can be set via --ui-config-file (#25132)

ServeurpersoCom committed 115 days ago

Verified 66589253

ggml : fix broken CPU concat implementation for quantized types (#25247)

fairydreaming committed 115 days ago

Verified ef2d7701

chat: trim messages sent to StepFun parser (fixes long reasoning loops) (#25238)

pwilkin committed 116 days ago

Verified 2d973636

ui: Improve performance when streaming (#25225)

ntowle committed 116 days ago

Verified d4cff114

ui: strip path and weight extension from model id in single model mode (#25137)

ServeurpersoCom committed 116 days ago

Verified f113e02d

spec: support spec-draft-p-min in DFlash (#25246)

ruixiang63 committed 116 days ago

Verified 152d337f

cuda: enable topk-moe fusion for 288 experts (#25267)

pwilkin committed 116 days ago

Verified 75a48a90

ui: align persisted config with strict server schema and enable thinking by default (#25242)

ServeurpersoCom committed 116 days ago

Verified 067de937

server + ui: ping silent SSE streams every 1s and kick only after 3s so slow prefill never drops healthy connections (#25241)

ServeurpersoCom committed 116 days ago

Verified b5315e16

ui: Add MCP Servers Opt-In for first time visitors (#25239)

allozaur committed 117 days ago

Verified 94875285

Remove redundant CUDA copies after gated_delta_net. (#23940)

gaugarg-nv committed 117 days ago

Verified 5a460dea

vendor : update cpp-httplib to 0.49.0 (#25218)

cabelo committed 117 days ago

Verified c8ae9a75

llama : add llama_model_ftype_name() (#25134)

angt committed 117 days ago

Verified fdb1db87

opencl: allow loading precompiled binary kernels from library (#23042)

lhez committed 118 days ago

Verified 4fc4ec55

common : use hf primary split as model path (#25194)

angt committed 118 days ago

Verified a6647b1a

Older