python

Workflow runs for python

• python

  completed successfully:
  python#18397:Merge group checks requested
  Oct 15, 20261m 36s
  More actions for workflow run 18397

• Fix out-of-bounds row access in inference layer_norm/rms_norm kernels for hidden sizes <= 128

  requires action with the application:
  python#18396:Pull request#8830openedbyjayden1711
  jayden1711:fix/inference-norm-row-guard

  Oct 15, 2026Action required
  More actions for workflow run 18396

• python

  completed successfully:
  python#18395:Merge group checks requested
  Oct 15, 20261m 39s
  More actions for workflow run 18395

• AutoEP + AutoTP folding: allow CPU optimizer offload under ZeRO-3

  completed successfully:
  python#18394:Pull request#8829openedby0z5a
  0z5a:autoep-folding-zero3-offload

  Oct 15, 20261m 48s
  More actions for workflow run 18394

• Add opt-in TorchAO HiFloat8 Linear and AutoEP training

  requires action with the application:
  python#18393:Pull request#8784synchronizebyhazelduan
  hazelduan:deepspeed_ao

  Oct 15, 2026Action required
  More actions for workflow run 18393

• Fix inference layer_norm/rms_norm launching no kernel for fp32 hidden sizes 65-128

  requires action with the application:
  python#18392:Pull request#8827synchronizebyjayden1711
  jayden1711:fix/inference-norm-threads-per-group-32

  Oct 15, 2026Action required
  More actions for workflow run 18392

• Fix free block handling in Inference V2 scheduling

  requires action with the application:
  python#18391:Pull request#8503synchronizebysj0618
  sj0618:fix/inference-v2-free-blocks

  Oct 15, 2026Action required
  More actions for workflow run 18391

• Fix inference layer_norm/rms_norm launching no kernel for fp32 hidden sizes 65-128

  requires action with the application:
  python#18390:Pull request#8827openedbyjayden1711
  jayden1711:fix/inference-norm-threads-per-group-32

  Oct 15, 2026Action required
  More actions for workflow run 18390

• [AutoSP] Aggregate the causal LM loss and gradients over the SP group

  requires action with the application:
  python#18389:Pull request#8457synchronizebyjinyouzhi
  jinyouzhi:autosp-correctness

  Oct 15, 2026Action required
  More actions for workflow run 18389

• Add opt-in TorchAO HiFloat8 Linear and AutoEP training

  requires action with the application:
  python#18388:Pull request#8784synchronizebyhazelduan
  hazelduan:deepspeed_ao

  Oct 15, 2026Action required
  More actions for workflow run 18388

• python

  completed successfully:
  python#18387:Merge group checks requested
  Oct 15, 20261m 36s
  More actions for workflow run 18387

• python

  completed successfully:
  python#18386:Merge group checks requested
  Oct 15, 20261m 36s
  More actions for workflow run 18386

• feat(rollout): segment-based kernel injection for hybrid-engine greedy decode

  completed successfully:
  python#18385:Pull request#8602synchronizebydelock
  gma/qwen_ki_expr

  Oct 15, 20261m 30s
  More actions for workflow run 18385

• Describe the CodeGen fused-QKV layout as a block selection

  completed successfully:
  python#18384:Pull request#8706synchronizebyAchyuthan-S
  Achyuthan-S:step5-codegen-affine-map

  Oct 15, 20261m 36s
  More actions for workflow run 18384

• python

  completed successfully:
  python#18383:Merge group checks requested
  Oct 15, 20261m 32s
  More actions for workflow run 18383

• python

  completed successfully:
  python#18382:Scheduled
  master

  Oct 15, 20261m 21s
  More actions for workflow run 18382

• feat(rollout): segment-based kernel injection for hybrid-engine greedy decode

  completed successfully:
  python#18381:Pull request#8602synchronizebydelock
  gma/qwen_ki_expr

  Oct 14, 20261m 29s
  More actions for workflow run 18381

• [CI] Run torch-latest only on AWS

  completed successfully:
  python#18380:Pull request#8825synchronizebytohtana
  tohtana:tohtana/torch-latest-aws-only

  Oct 14, 20261m 42s
  More actions for workflow run 18380

• [CI] Run torch-latest only on AWS

  completed successfully:
  python#18379:Pull request#8825openedbytohtana
  tohtana:tohtana/torch-latest-aws-only

  Oct 14, 20261m 33s
  More actions for workflow run 18379

• Reduce chunked cross-entropy memory test footprint

  completed successfully:
  python#18378:Pull request#8824synchronizebytohtana
  tohtana:tohtana/reduce-cross-entropy-test-memory

  Oct 14, 20261m 36s
  More actions for workflow run 18378

• Reduce chunked cross-entropy memory test footprint

  completed successfully:
  python#18377:Pull request#8824openedbytohtana
  tohtana:tohtana/reduce-cross-entropy-test-memory

  Oct 14, 20261m 52s
  More actions for workflow run 18377

• Fix routed-expert gradient clipping for AutoEP + AutoTP folding

  requires action with the application:
  python#18376:Pull request#8823synchronizeby0z5a
  0z5a:codex/autoep-active-clipping-20261012

  Oct 14, 2026Action required
  More actions for workflow run 18376

• python

  completed successfully:
  python#18375:Merge group checks requested
  Oct 14, 20261m 37s
  More actions for workflow run 18375

• Fix routed-expert gradient clipping for AutoEP + AutoTP folding

  requires action with the application:
  python#18374:Pull request#8823openedby0z5a
  0z5a:codex/autoep-active-clipping-20261012

  Oct 14, 2026Action required
  More actions for workflow run 18374

• Add opt-in TorchAO HiFloat8 Linear and AutoEP training

  requires action with the application:
  python#18373:Pull request#8784synchronizebyhazelduan
  hazelduan:deepspeed_ao

  Oct 14, 2026Action required
  More actions for workflow run 18373