Core AI

Workflow runs for Core AI

• PDX fast path for spatial mean.out of channels-last tensors (#24164)

  completed successfully:
  Core AI#655:Commite875b0fpushedbyrezaasjd
  main

  Oct 29, 202644s
  More actions for workflow run 655

• XNNPACKQuantizer: apply module filters to ops only, not weight nodes …

  completed successfully:
  Core AI#654:Commit10b5018pushedbynil-is-all
  main

  Oct 29, 202630s
  More actions for workflow run 654

• Vulkan: keep a reshape that follows a dynamically quantized linear (#…

  completed successfully:
  Core AI#653:Commitd15a629pushedbynil-is-all
  main

  Oct 29, 202626m 45s
  More actions for workflow run 653

• Run several ExecuTorch CUDA test targets on A100 MIG instead of exclu…

  completed successfully:
  Core AI#652:Commitc9fba66pushedbymeta-codesync[bot]Bot
  main

  Oct 29, 202636s
  More actions for workflow run 652

• Normalize permute dims in PostponePermuteOpBelowSqueezeOrUnsqueezeLik…

  completed successfully:
  Core AI#651:Commit68a080apushedbymeta-codesync[bot]Bot
  main

  Oct 29, 202639s
  More actions for workflow run 651

• Add skip_h2d_for_constants to keep constants in device memory (#24274)

  completed successfully:
  Core AI#650:Commitb80295bpushedbyJacobSzwejbka
  release/1.6

  Oct 29, 202625m 5s
  More actions for workflow run 650

• Qualcomm AI Engine Direct - Keep a non-delegated op's weight dequanti…

  completed successfully:
  Core AI#649:Commit2cb048epushedbyJacobSzwejbka
  release/1.6

  Oct 29, 202632s
  More actions for workflow run 649

• exir: Track unbacked symbols during pass replay (#24263)

  completed successfully:
  Core AI#648:Commit0842bfapushedbyJacobSzwejbka
  release/1.6

  Oct 29, 202630s
  More actions for workflow run 648

• Test optimized ESP-NN kernels on the S3 emulator (#24244)

  completed successfully:
  Core AI#647:Commit043c65bpushedbyrascani
  main

  Oct 29, 202642s
  More actions for workflow run 647

• Filter py_version from OSS Buck rule arguments (#24308)

  completed successfully:
  Core AI#646:Commitd4bf403pushedbyJakeStevens
  main

  Oct 29, 202622m 43s
  More actions for workflow run 646

• Espressif: Share one ESP-IDF PAL across examples (#24291)

  completed successfully:
  Core AI#645:Commit6bb588cpushedbyJakeStevens
  main

  Oct 29, 202639s
  More actions for workflow run 645

• Arm backend: Reland FP8 VGF operator tests (#24296)

  completed successfully:
  Core AI#644:Commitcd461b0pushedbyJakeStevens
  main

  Oct 29, 202641s
  More actions for workflow run 644

• [XNNPACK] Lower fp32 <-> fp16 casts as XNNConvert (#24277)

  completed successfully:
  Core AI#643:Commit9e3ec43pushedbyJakeStevens
  main

  Oct 29, 202644s
  More actions for workflow run 643

• Package Native, Vulkan, and GGML as an executorch-native wheel (#24…

  completed successfully:
  Core AI#642:Commita0a365bpushedbymeta-codesync[bot]Bot
  main

  Oct 29, 202628m 7s
  More actions for workflow run 642

• Arm backend: Reland tensorized GPT-OSS MoE regression tests (#24297)

  completed successfully:
  Core AI#641:Commit724da85pushedbyJakeStevens
  main

  Oct 29, 202642s
  More actions for workflow run 641

• Remove redundant join ownership bookkeeping from batching shutdown

  completed successfully:
  Core AI#640:Commit84c701bpushedbymeta-codesync[bot]Bot
  main

  Oct 29, 202639s
  More actions for workflow run 640

• Make Neuron adapter selection version-aware (#24160) (#24160)

  completed successfully:
  Core AI#639:Commit14b5e1dpushedbypsiddh
  main

  Oct 29, 20261m 1s
  More actions for workflow run 639

• Make FCB lowering and cleanup safe (#24159) (#24159)

  completed successfully:
  Core AI#638:Commit012fd14pushedbypsiddh
  main

  Oct 29, 202629m 32s
  More actions for workflow run 638

• [XNNPACK] Fold zero constant_pad_nd into convolutions (#24112)

  completed successfully:
  Core AI#637:Commit810f2fdpushedbyJakeStevens
  main

  Oct 29, 202649s
  More actions for workflow run 637

• Add a Native CPU engine backed by CPUPlan (#23935)

  completed successfully:
  Core AI#636:Commit1d55f9cpushedbymeta-codesync[bot]Bot
  main

  Oct 29, 202641s
  More actions for workflow run 636

• Arm backend: Prepare TOSA validation and lowering for int64 (#24128)

  completed successfully:
  Core AI#635:Commit6a2784epushedbyzingo
  main

  Oct 29, 202642s
  More actions for workflow run 635

• Arm backend: Fix Embedding provenenance. (#24280)

  completed successfully:
  Core AI#634:Commit237f31fpushedbywwwind
  main

  Oct 29, 202620m 45s
  More actions for workflow run 634

• Arm backend: Add static public API manifest for 1.6 (#24290)

  completed successfully:
  Core AI#633:Commit5c1112cpushedbySebastian-Larsson
  main

  Oct 29, 202657s
  More actions for workflow run 633

• Support 1D linear in the CPU Reference provider

  completed successfully:
  Core AI#632:Commit22df868pushedbymeta-codesync[bot]Bot
  main

  Oct 29, 202656s
  More actions for workflow run 632

• Arm backend: Decompose mv and dot operators before quantization (#24284)

  completed successfully:
  Core AI#631:Commit7c72b5dpushedbyErik-Lundell
  main

  Oct 29, 202638s
  More actions for workflow run 631