All pull requests

Pull requests

• Grid-stride FP8 fixed-NK group setup (#6481)

  cla signedmeta-exported
  #6481·q10 opened Oct 16, 2026Contributor·

  ·
  1

• Replace legacy null CUDA stream constants

  cla signedmeta-exported
  #6479·q10 opened Oct 16, 2026Contributor·

  ·
  1

• Speed up fbgemm transpose_avx512 tail by replacing cross-lane shuffles with permutes

  cla signedmeta-exported
  #6478·meta-codesync[bot] opened Oct 15, 2026Bot·

  ·

• Guard quantizer devices and handle zero-width rows

  cla signedmeta-exported
  #6477·q10 opened Oct 15, 2026Contributor·

  ·
  1

• Honor FP8 matmul output dtype for empty inputs (#6476)

  cla signedmeta-exported
  #6476·han3101 opened Oct 15, 2026Contributor·

  ·
  1

• Rate-limit the CPU bounds_check_indices out-of-range warning log (#6475)

  cla signedmeta-exported
  #6475·kmonian opened Oct 15, 2026·

  ·
  1

• Move the ROCm CI to ROCm 7.14 and 10.0 (#6474)

  ciflow/rocmcla signedmeta-exportedmodule: rocm
  #6474·q10 opened Oct 15, 2026Contributor·

  ·
  2

• Avoid eager empty tensors in INT-N-bit inference dispatch (#6473)

  cla signedmeta-exported
  #6473·mrmiywj opened Oct 15, 2026Contributor·

  ·
  1

• Re-enable TBE backward none and SGD writeback tests on ROCm (#6472)

  ciflow/rocmcla signedmeta-exportedmodule: rocm
  #6472·q10 opened Oct 15, 2026Contributor·

  ·
  2

• Suppress type errors for Pyre upgrade (landing group 36)

  cla signedmeta-exported
  #6471·maggiemoss opened Oct 15, 2026Contributor·

  ·
  1

• Match FP8 matmul fake shapes to eager ops (#6470)

  cla signedmeta-exported
  #6470·han3101 opened Oct 15, 2026Contributor·

  ·
  1

• Migrate codegen inference mode to constraints

  cla signedmeta-exported
  #6469·scottliu77 opened Oct 14, 2026·

  ·

• Shard FBGEMM backward warp codegen by grad dtype (#6468)

  cla signedmeta-exported
  #6468·kainanpeace666 opened Oct 14, 2026·

  ·
  1

• Detect gfx1250 and select TheRock CI mode

  cla signedmeta-exported
  #6463·q10 opened Oct 14, 2026Contributor·

  ·
  1

• Honor scale_bias_last in the N-bit dequantization entry point

  cla signed
  #6461·cyyever opened Oct 14, 2026Contributor·

  ·
  1

• Template SVE N-bit input widths

  cla signed
  #6460·cyyever opened Oct 14, 2026Contributor·

  ·
  1

• Remove shadowing from standalone GPU operators

  cla signedmeta-exported
  #6457·q10 opened Oct 13, 2026Contributor·

  ·
  1

• Fix TBE forward using stale uvm_cache_stats

  cla signed
  #6454·avbokovoy opened Oct 13, 2026Contributor·

  ·
  1

• Re-enable TBE backward none and SGD writeback tests on ROCm

  ciflow/rocmcla signedmodule: rocm
  #6451·avbokovoy opened Oct 13, 2026Contributor·

  ·
  2

• Name split-K shared-memory sizes by role

  cla signedmeta-exported
  #6450·q10 opened Oct 13, 2026Contributor·

  ·
  1

• Guard gfx942 INT4 no-bag specialization (#6446)

  cla signedmeta-exported
  #6446·mrmiywj opened Oct 12, 2026Contributor·

  ·
  1

• Parallelize gfx942 INT4 D64 no-bag lookup (#6445)

  cla signedmeta-exported
  #6445·mrmiywj opened Oct 12, 2026Contributor·

  ·
  1

• Mark invalid cache lookup locations as missing

  cla signedmeta-exported
  #6436·q10 opened Oct 11, 2026Contributor·

  ·
  1

• Pass complete VBE metadata to SSD bounds checks (#6435)

  cla signedmeta-exported
  #6435·q10 opened Oct 10, 2026Contributor·

  ·
  1

• Suppress type errors for Pyre upgrade (landing group 21)

  cla signedmeta-exported
  #6429·maggiemoss opened Oct 9, 2026Contributor·

  ·
  1