ggerganov/llama.cpp

Pull Requests Commits

contrib : rewrite AGENTS.md, make it more clear about project values (#21270)

ngxson committed 8 days ago

Verified c30e0122

opencl: fix leak in Adreno q8_0 path (#21212)

lhez committed 8 days ago

Verified 95a6ebab

server: Bypass API Key validation for WebUI static bundle assets (#21269)

allozaur committed 8 days ago

Verified 12dbf1da

CUDA: fix FA kernel selection logic (#21271)

JohannesGaessler committed 8 days ago

Verified 86221cf6

kleidiai: add CPU feature detection to CI run script (#20394)

martin-klacer-arm committed 9 days ago

Verified 6de97b9d

Update Dawn version in WebGPU CI (#20784)

nikhilJain17 committed 9 days ago

Verified 5a0ed515

hexagon: improve RMS_NORM and DIV accuracy (#21251)

aparmp-quic committed 9 days ago

Verified 8710e5f9

fix: tool call parsing for LFM2 and LFM2.5 models (#21242)

jbuchananr committed 9 days ago

Verified 1d6d4cf7

llama : rotate activations for better quantization (#21038)

ggerganov committed 9 days ago

Verified 744c0c73

scripts: add function call test script (#21234)

ngxson committed 9 days ago

Verified 0356e33a

ggerganov committed 9 days ago

6422036f

ggml : bump version to 0.9.10 (ggml/1454)

ggerganov committed 9 days ago

296bc053

sycl : support nvfp4 type in mul_mat (#21227)

arthw committed 9 days ago

Verified 6b949d10

ggml-cuda: Add generic NVFP4 MMQ kernel (#21074)

michaelw9999 committed 9 days ago

Verified 84f82e84

memory: respect unified KV cache in hybrid memory for eval tasks (#21224)

mudler committed 9 days ago

Verified e1cb8174

CUDA/HIP: Fix kernel slection for mmvq mmid kernel to align host selection with device launch bounds (#21238)

IMbackK committed 9 days ago

Verified 88d5f8ff

ggml : fix RWKV ops thread assignment (#21226)

ggerganov committed 9 days ago

Verified d43375ff

ggml-cpu: fix fallback for RVV kernels without zvfh (#21157)

taimur-10x committed 9 days ago

Verified 2b86e5ca

CUDA: Add Flash Attention Support for Head Dimension 512 (#20998)

anavp-nvidia committed 9 days ago

Verified 88458164

llama : refactor llama_model_quantize_params to expose a pure C interface (#20346)

EAddario committed 9 days ago

Verified 49512502

ggml webgpu: quantized buffers to u32 + wider browser/device support (#21046)

reeselevine committed 9 days ago

Verified 82764c34

ggml-webgpu: port all AOT operators to JIT (#20728)

abhijitramesh committed 9 days ago

Verified 825eb91a

fix: Use lower-case proxy headers naming (#21235)

allozaur committed 10 days ago

Verified 0fcb3760

common : cleanup logs and modernize the progress bar (#21215)

angt committed 10 days ago

Verified 6307ec07

CANN: fix multi-thread set_tensor race conditions (#20151)

hipudding committed 10 days ago

Verified 632219af

server: (webui) no more gzip compression (#21073)

ngxson committed 10 days ago

Verified 4a00bbfe

common : gpt-oss handle builtin and unsolicited tool calls (#21213)

aldehir committed 10 days ago

Verified 624733d6

fix: correct misspellings in code comments (#21217)

lainon1 committed 10 days ago

Verified 0b6ff479

CI: Enable CPU and Vulkan ARM64 Release (#21207)

ehfd committed 10 days ago

Verified eec6f85d

ggerganov committed 10 days ago

9281dd13

Older