Claude Code

Workflow runs for Claude Code

• Support 3-D convolution in the portable convolution kernel

  skipped:
  Claude Code#18753:Issue comment#24385 (comment)createdbyclaude[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18753

• Support 3-D convolution in the portable convolution kernel

  currently running:
  Claude Code#18752:Issue comment#24385 (comment)createdbyshoumikhin
  Oct 31, 2026In progress
  More actions for workflow run 18752

• Arm backend: Quantize argmax input so float logits lower to TOSA ARGMAX

  skipped:
  Claude Code#18751:Issue comment#24394 (comment)createdbymeta-cla[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18751

• Arm backend: Quantize argmax input so float logits lower to TOSA ARGMAX

  skipped:
  Claude Code#18750:Issue comment#24394 (comment)createdbypcwilk
  Oct 31, 20261s
  More actions for workflow run 18750

• Qualcomm AI Engine Direct - Keep node names unique across the QNN and Samsung passes, and decompose constant-only ops on HTP

  skipped:
  Claude Code#18749:Issue comment#24316 (comment)createdbyshoumikhin
  Oct 31, 20261s
  More actions for workflow run 18749

• Qualcomm AI Engine Direct - Keep node names unique across the QNN and Samsung passes, and decompose constant-only ops on HTP

  skipped:
  Claude Code#18748:Issue comment#24316 (comment)createdbyshoumikhin
  Oct 31, 20261s
  More actions for workflow run 18748

• Support 3-D convolution in the portable convolution kernel

  skipped:
  Claude Code#18747:Issue comment#24385 (comment)createdbyshoumikhin
  Oct 31, 20261s
  More actions for workflow run 18747

• Arm backend: Quantize argmax input so float logits lower to TOSA ARGMAX

  skipped:
  Claude Code#18746:Issue comment#24394 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18746

• Arm backend: Quantize argmax input so float logits lower to TOSA ARGMAX

  skipped:
  Claude Code#18745:Issue comment#24394 (comment)createdbylinux-foundation-easycla[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18745

• Arm backend: Quantize argmax input so float logits lower to TOSA ARGMAX

  skipped:
  Claude Code#18744:Issue comment#24394 (comment)createdbymeta-cla[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18744

• Arm backend: Quantize argmax input so float logits lower to TOSA ARGMAX

  skipped:
  Claude Code#18743:Issue comment#24394 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 20267s
  More actions for workflow run 18743

• Qualcomm AI Engine Direct - Keep node names unique across the QNN and Samsung passes, and decompose constant-only ops on HTP

  skipped:
  Claude Code#18742:Issue comment#24316 (comment)createdbyclaude[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18742

• Qualcomm AI Engine Direct - Keep node names unique across the QNN and Samsung passes, and decompose constant-only ops on HTP

  completed successfully:
  Claude Code#18741:Issue comment#24316 (comment)createdbyshoumikhin
  Oct 31, 202631m 42s
  More actions for workflow run 18741

• Fix mixed-dtype cat and stack in to_edge, the edge check and portable cat

  skipped:
  Claude Code#18740:Issue comment#24381 (comment)createdbyshoumikhin
  Oct 31, 20261s
  More actions for workflow run 18740

• Python runtime: write runtime logs straight to stderr so two threads cannot deadlock

  skipped:
  Claude Code#18739:Issue comment#24393 (comment)createdbyclaude[bot]Bot
  Oct 31, 20268s
  More actions for workflow run 18739

• Python runtime: write runtime logs straight to stderr so two threads cannot deadlock

  completed successfully:
  Claude Code#18738:Issue comment#24393 (comment)createdbyshoumikhin
  Oct 31, 202621m 54s
  More actions for workflow run 18738

• Python runtime: write runtime logs straight to stderr so two threads cannot deadlock

  skipped:
  Claude Code#18737:Issue comment#24393 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18737

• [ET-VK][benchmark] Add JSON operator and shader benchmark cases

  skipped:
  Claude Code#18736:Issue comment#24391 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 2026Skipped
  More actions for workflow run 18736

• [ET-VK][sdpa] Expose shader override

  skipped:
  Claude Code#18735:Issue comment#24390 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18735

• [ET-VK][runtime] Expose raw shader timestamps and dispatch arguments

  skipped:
  Claude Code#18734:Issue comment#24389 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18734

• [ET-VK][runtime] Support arbitrary explicit local workgroup sizes

  skipped:
  Claude Code#18733:Issue comment#24388 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18733

• [ET-VK][runtime] Derive SPIR-V requirements and enable Vulkan shader features

  skipped:
  Claude Code#18732:Issue comment#24387 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18732

• [ET-VK][q8ta] Honor alpha in quantized addition

  skipped:
  Claude Code#18731:Issue comment#24386 (comment)createdbypytorch-bot[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18731

• Fix 0-d tensors, dynamic output sizes and integer reductions, and default to float32 on CPU, in the OpenVINO backend

  skipped:
  Claude Code#18730:Issue comment#24377 (comment)createdbyclaude[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18730

• Python runtime: enforce CUDA stream selection rules on the stream itself

  skipped:
  Claude Code#18729:Issue comment#24346 (comment)createdbyclaude[bot]Bot
  Oct 31, 20261s
  More actions for workflow run 18729