Search results

vulkan: matmul dequantization improvements

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 12015

#12015 In ggml-org/llama.cpp;

· netrunnereve opened on Feb 21, 2025

vulkan: implement more backpropagation operators

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 11914

#11914 In ggml-org/llama.cpp;

· remyoudompheng opened on Feb 16, 2025

Upgrade init_tensor API to return a ggml_status

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

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

testingEverything test relatedEverything test related

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open (in progress).

ggml-org/llama.cppnumber 11854

#11854 In ggml-org/llama.cpp;

· WilliamTambellini opened on Feb 14, 2025

vulkan: improve im2col and RDNA1 performance

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 11826

#11826 In ggml-org/llama.cpp;

· daniandtheweb opened on Feb 12, 2025

vulkan: add specific MMV kernels for IQ2 and IQ3 quants + optimizations

devopsimprovements to build systems and github actionsimprovements to build systems and github actions

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 11595

#11595 In ggml-org/llama.cpp;

· remyoudompheng opened on Feb 2, 2025

llama: Add support for RWKV v7 architecture

Apple Metalhttps://en.wikipedia.org/wiki/Metal_(API)https://en.wikipedia.org/wiki/Metal_(API)

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

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

testingEverything test relatedEverything test related

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open (in progress).

ggml-org/llama.cppnumber 11452

#11452 In ggml-org/llama.cpp;

· MollySophia opened on Jan 27, 2025

vulkan : add GGML_VK_FORCE_HEAP_INDEX env var

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 9734

#9734 In ggml-org/llama.cpp;

· gyf304 opened on Oct 4, 2024

vulkan: fix storageBuffer16BitAccess detection on some adreno driver

bugfixfixes an issue or bugfixes an issue or bug

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

Review Complexity : LowTrivial changes to code that most beginner devs (or those who want a break) can tackle. e.g. UI fixTrivial changes to code that most beginner devs (or those who want a break) can tackle. e.g. UI fix

VulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backend

Status: Open (in progress).

ggml-org/llama.cppnumber 8837

#8837 In ggml-org/llama.cpp;

· rhjdvsgsgks opened on Aug 2, 2024