Search results

SYCL: Add set_rows support for quantized types

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

SYCLhttps://en.wikipedia.org/wiki/SYCL - GPU programming languagehttps://en.wikipedia.org/wiki/SYCL - GPU programming language

Status: Open (in progress).

ggml-org/llama.cppnumber 14883

#14883 In ggml-org/llama.cpp;

· qnixsynapse opened on Jul 26, 2025

GGML: Fix leak of backend buffer memory address in RPC

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

Status: Open (in progress).

ggml-org/llama.cppnumber 14882

#14882 In ggml-org/llama.cpp;

· struct opened on Jul 26, 2025

sync : ggml

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

scriptScript relatedScript related

Status: Open (in progress).

ggml-org/llama.cppnumber 14873

#14873 In ggml-org/llama.cpp;

· ggerganov opened on Jul 25, 2025

vulkan : add fp16 support for the conv_2d kernel

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

testingEverything test relatedEverything test related

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open (in progress).

ggml-org/llama.cppnumber 14872

#14872 In ggml-org/llama.cpp;

· Green-Sky opened on Jul 25, 2025

vulkan: skip empty set_rows to avoid invalid API usage

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open (in progress).

ggml-org/llama.cppnumber 14860

#14860 In ggml-org/llama.cpp;

· jeffbolznv opened on Jul 24, 2025

test-backend-ops: enables perf/eval testing of composite ops

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

testingEverything test relatedEverything test related

Status: Open (in progress).

ggml-org/llama.cppnumber 14833

#14833 In ggml-org/llama.cpp;

· etasnadi opened on Jul 23, 2025

sycl: refactor quantization to q8_1

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

SYCLhttps://en.wikipedia.org/wiki/SYCL - GPU programming languagehttps://en.wikipedia.org/wiki/SYCL - GPU programming language

Status: Open (in progress).

ggml-org/llama.cppnumber 14815

#14815 In ggml-org/llama.cpp;

· Alcpz opened on Jul 22, 2025

opencl: tiled mul_mat with local memory for f16 and f32

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

OpenCLIssues specific to the OpenCL backendIssues specific to the OpenCL backend

Status: Draft (not ready).

ggml-org/llama.cppnumber 14809

#14809 In ggml-org/llama.cpp;

· lhez opened on Jul 22, 2025

CUDA: skip masked out KQ slices in mma FA kernel

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

Nvidia GPUIssues specific to Nvidia GPUsIssues specific to Nvidia GPUs

pythonpython script changespython script changes

Status: Open (in progress).

ggml-org/llama.cppnumber 14735

#14735 In ggml-org/llama.cpp;

· JohannesGaessler opened on Jul 17, 2025

Fix KleidiAI compilation errors with -DGGML_NATIVE=OFF (issue #14464)

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

Status: Open (in progress).

ggml-org/llama.cppnumber 14700

#14700 In ggml-org/llama.cpp;

· baonudesifeizhai opened on Jul 15, 2025

Add Pad Reflect 1D CUDA support

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

Nvidia GPUIssues specific to Nvidia GPUsIssues specific to Nvidia GPUs

Status: Open (in progress).

ggml-org/llama.cppnumber 14659

#14659 In ggml-org/llama.cpp;

· YavorGIvanov opened on Jul 13, 2025

Add CUDA non-contiguous Unary Ops support

buildCompilation issuesCompilation issues

documentationImprovements or additions to documentationImprovements or additions to documentation

ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learning

Nvidia GPUIssues specific to Nvidia GPUsIssues specific to Nvidia GPUs

testingEverything test relatedEverything test related

Status: Open (in progress).

ggml-org/llama.cppnumber 14639

#14639 In ggml-org/llama.cpp;

· YavorGIvanov opened on Jul 11, 2025