Issues
state:open label:Vulkan
state:open label:Vulkan
Search results
ggml : support broadcast for ggml_soft_max_ext and ggml_flash_attn_ext
Apple Metalhttps://en.wikipedia.org/wiki/Metal_(API)https://en.wikipedia.org/wiki/Metal_(API)Ascend NPUissues specific to Ascend NPUsissues specific to Ascend NPUsggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningNvidia GPUIssues specific to Nvidia GPUsIssues specific to Nvidia GPUsSYCLhttps://en.wikipedia.org/wiki/SYCL - GPU programming languagehttps://en.wikipedia.org/wiki/SYCL - GPU programming languagetestingEverything test relatedEverything test relatedVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 14435#14435 In ggml-org/llama.cpp;vulkan: Add fusion support for RMS_NORM+MUL
ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningtestingEverything test relatedEverything test relatedVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 14366#14366 In ggml-org/llama.cpp;ggml: adds CONV_2D op and direct GEMM Vulkan implementation
ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningtestingEverything test relatedEverything test relatedVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 14316#14316 In ggml-org/llama.cpp;ggml : implement REGLU/GEGLU/SWIGLU ops
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 learninghelp wantedExtra attention is neededExtra attention is neededNvidia GPUIssues specific to Nvidia GPUsIssues specific to Nvidia GPUsSYCLhttps://en.wikipedia.org/wiki/SYCL - GPU programming languagehttps://en.wikipedia.org/wiki/SYCL - GPU programming languagetestingEverything test relatedEverything test relatedVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Draft (not ready).ggml-org/llama.cppnumber 14158#14158 In ggml-org/llama.cpp;Fix Vulkan glslc invocation command lines
ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 13289#13289 In ggml-org/llama.cpp;llama : try loading tensors with pre-computed hashes
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 learningKomputehttps://github.com/KomputeProject/kompute/https://github.com/KomputeProject/kompute/Nvidia GPUIssues specific to Nvidia GPUsIssues specific to Nvidia GPUsSYCLhttps://en.wikipedia.org/wiki/SYCL - GPU programming languagehttps://en.wikipedia.org/wiki/SYCL - GPU programming languageVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 13106#13106 In ggml-org/llama.cpp;llama-server : implement universal assisted decoding
androidIssues specific to AndroidIssues specific to AndroidApple Metalhttps://en.wikipedia.org/wiki/Metal_(API)https://en.wikipedia.org/wiki/Metal_(API)Ascend NPUissues specific to Ascend NPUsissues specific to Ascend NPUsbuildCompilation issuesCompilation issuesdevopsimprovements to build systems and github actionsimprovements to build systems and github actionsdocumentationImprovements or additions to documentationImprovements or additions to documentationggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningNvidia GPUIssues specific to Nvidia GPUsIssues specific to Nvidia GPUspythonpython script changespython script changesscriptScript relatedScript relatedSYCLhttps://en.wikipedia.org/wiki/SYCL - GPU programming languagehttps://en.wikipedia.org/wiki/SYCL - GPU programming languagetestingEverything test relatedEverything test relatedVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 12635#12635 In ggml-org/llama.cpp;Vulkan: Remove dedicated aligned matrix matrix multiplication shaders
ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningtestingEverything test relatedEverything test relatedVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Draft (not ready).ggml-org/llama.cppnumber 12515#12515 In ggml-org/llama.cpp;Fixed Eval Bug: 12163 : Fallback to CPU when loading model: vk::PhysicalDevice::createDevice: ErrorExtensionNotPresent.
ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 12329#12329 In ggml-org/llama.cpp;<code>tool-call</code>: Phi-4 support
androidIssues specific to AndroidIssues specific to AndroidApple Metalhttps://en.wikipedia.org/wiki/Metal_(API)https://en.wikipedia.org/wiki/Metal_(API)devopsimprovements to build systems and github actionsimprovements to build systems and github actionsdocumentationImprovements or additions to documentationImprovements or additions to documentationggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningNvidia GPUIssues specific to Nvidia GPUsIssues specific to Nvidia GPUspythonpython script changespython script changesSYCLhttps://en.wikipedia.org/wiki/SYCL - GPU programming languagehttps://en.wikipedia.org/wiki/SYCL - GPU programming languagetestingEverything test relatedEverything test relatedVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 12288#12288 In ggml-org/llama.cpp;vulkan: optimization proposals for coopmat1 mul_mm
ggmlchanges relating to the ggml tensor library for machine learningchanges relating to the ggml tensor library for machine learningVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Draft (not ready).ggml-org/llama.cppnumber 12260#12260 In ggml-org/llama.cpp;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 learningVulkanIssues specific to the Vulkan backendIssues specific to the Vulkan backendStatus: Open (in progress).ggml-org/llama.cppnumber 9734#9734 In ggml-org/llama.cpp;