All pull requests

Pull requests

• riscv: add bf16 SELU inference

  riscv
  #7196·luke-opus opened Oct 29, 2026Contributor·

  ·
  3

• riscv: add bf16 Eltwise inference

  riscv
  #7195·luke-opus opened Oct 29, 2026Contributor·

  ·
  3

• riscv: add bf16 BatchNorm inference

  riscv
  #7192·luke-opus opened Oct 28, 2026Contributor·

  ·
  3

• riscv: add bf16 PReLU inference

  riscv
  #7189·luke-opus opened Oct 27, 2026Contributor·

  ·
  3

• fix sdpa cross attention reusing the kvcache

  armdoclayerloongarchmipsriscvtestvulkanx86
  #7188·futz12 opened Oct 27, 2026Contributor·

  ·
  1

• fix multiheadattention cross attention reusing its kvcache

  armlayerloongarchmipsriscvtestvulkanx86
  #7187·futz12 opened Oct 27, 2026Contributor·

  ·
  3

• pnnx: honor fp16=False for shape-only Python conversion

  pnnxtool
  #7184·ktz03 opened Oct 27, 2026Contributor·

  ·
  1

  1

• pnnx: stabilize static Float32 LogSoftmax conversion

  pnnxtool
  #7178·ktz03 opened Oct 26, 2026Contributor·

  ·
  1

• pnnx: preserve tanh mode in fused GELU formulas

  pnnxtool
  #7176·ktz03 opened Oct 26, 2026Contributor·

  ·
  1

  1

• vulkan: avoid fp16 overflow in Normalize coefficients

  testvulkan
  #7174·kongxinjun opened Oct 25, 2026Contributor·

  ·
  1

• pnnx: preserve complex scalar expression precision

  pnnxtool
  #7169·ktz03 opened Oct 24, 2026Contributor·

  ·
  1

  1

• Vulkan paged weight allocator

  coredoctestvulkan
  #7167·nihui opened Oct 23, 2026Member·

  ·
  3

• feat: use reservoir sampling for LLM calibration

  testtool
  #7166·Edwardssss opened Oct 23, 2026Contributor·

  ·
  4

• pnnx: support nearest-exact interpolation on x86 CPU

  docpnnxtesttoolx86
  #7163·Edwardssss opened Oct 22, 2026Contributor·

  ·
  6

• riscv: add bf16 InnerProduct inference

  riscv
  #7161·luke-opus opened Oct 21, 2026Contributor·

  ·
  3

• pnnx: escape python keywords in generated module attribute names

  pnnxtool
  #7159·drakeo338 opened Oct 21, 2026·

  ·
  1

  2

• pnnx: support positive-step tensor slicing with StridedSlice

  coredoclayerpnnxtesttool
  #7158·Edwardssss opened Oct 21, 2026Contributor·

  ·
  1

• riscv: vectorize fp32 InnerProduct for unpacked 2D inputs

  riscv
  #7155·luke-opus opened Oct 20, 2026Contributor·

  ·
  1

• Validate runtime blob shapes in GRU, RNN, LSTM, DeformableConv2D, Gemm and MultiHeadAttention, and check int8 scale records

  armlayerloongarchmipsriscvtestx86
  #7152·professor-moody opened Oct 20, 2026Contributor·

  ·
  7

• vulkan: add a barrier before record_clone writes the dst buffer

  core
  #7151·kongxinjun opened Oct 20, 2026Contributor·

  ·
  2

• fix integer overflow and missing dimension checks in yuv420sp pixel apis

  core
  #7149·ihb2032 opened Oct 20, 2026Contributor·2 tasks done·

  ·
  1

  1

• Validate multi-blob layer arity, ModelBin shape products, and layer weight/parameter invariants

  armcorelayerloongarchmipsriscvtestvulkanx86
  #7148·professor-moody opened Oct 19, 2026Contributor·

  ·
  17

• pnnx: add Caffe frontend

  pnnxtool
  #7146·nihui opened Oct 19, 2026Member·

  ·
  1

• vulkan: single pass square sum for Normalize on subgroups

  testvulkan
  #7143·futz12 opened Oct 19, 2026Contributor·

  ·
  3

• vulkan: reduce the flash attention row statistics on subgroups

  coretestvulkan
  #7142·futz12 opened Oct 19, 2026Contributor·

  ·
  4