Code Quality

Workflow runs for Code Quality

• [1/4] fix(speculative): four training fixes behind the Muse-Spark-Open DFlash2 recipe

  completed successfully:
  Code Quality#17705:Pull request#2617synchronizebyh-guo18
  haoguo/dflash-rope-fp32

  Oct 31, 20265m 26s
  More actions for workflow run 17705

• [OMNIML-6073] Quantize Nemotron-H's experts under grouped_mm, with their wrapper in its model folder

  completed successfully:
  Code Quality#17704:Pull request#3067synchronizebyshengliangxu
  shengliangx/hfep-8b-nemotron-h-experts

  Oct 31, 20264m 36s
  More actions for workflow run 17704

• [OMNIML-6071] Calibrate a distributed model on per-rank data

  completed successfully:
  Code Quality#17703:Pull request#2740synchronizebyshengliangxu
  shengliangx/hfep-5-distributed-calibration

  Oct 31, 20264m 40s
  More actions for workflow run 17703

• [OMNIML-6076] [2/3] Prepare a model's config for expert parallelism from its spec

  completed successfully:
  Code Quality#17702:Pull request#3135synchronizebyshengliangxu
  shengliangx/hfep-9a2-ep-spec-applier

  Oct 31, 20265m 9s
  More actions for workflow run 17702

• [OMNIML-6054] [1/2] Move fsdp2_wrap out of distributed.py into fsdp2_utils

  completed successfully:
  Code Quality#17701:Pull request#3140synchronizebyshengliangxu
  shengliangx/hfep-13a-move-fsdp2-wrap

  Oct 31, 20264m 58s
  More actions for workflow run 17701

• [OMNIML-6073] Quantize Nemotron-H's experts under grouped_mm, with their wrapper in its model folder

  completed successfully:
  Code Quality#17700:Pull request#3067synchronizebyshengliangxu
  shengliangx/hfep-8b-nemotron-h-experts

  Oct 31, 20265m 39s
  More actions for workflow run 17700

• Run fused-MoE expert GPTQ updates on all of this process's GPUs

  completed successfully:
  Code Quality#17699:Pull request#3002synchronizebycjluo-nv
  chenjiel/gptq-parallel-expert-updates

  Oct 31, 20265m 29s
  More actions for workflow run 17699

• hf_ptq: save/restore quantized state to retry export without recalibrating

  requires action with the application:
  Code Quality#17698:Pull request#2129synchronizebywyattearp
  wyattearp:ptq-reliability-fixes

  Oct 31, 2026Action required
  More actions for workflow run 17698

• [OMNIML-6273] Let the Claude review bot submit its review again

  completed successfully:
  Code Quality#17697:Pull request#3156openedbyshengliangxu
  shengliangx/claude-review-allowlist-fix

  Oct 31, 20266m 18s
  More actions for workflow run 17697

• testing doctest

  failed:
  Code Quality#17696:Pull request#2652synchronizebygrzegorz-k-karch
  gkarch/doctest

  Oct 31, 20265m 4s
  More actions for workflow run 17696

• [OMNIML-6071] Calibrate a distributed model on per-rank data

  completed successfully:
  Code Quality#17695:Pull request#2740synchronizebyshengliangxu
  shengliangx/hfep-5-distributed-calibration

  Oct 31, 20265m 10s
  More actions for workflow run 17695

• [OMNIML-6076] [2/3] Prepare a model's config for expert parallelism from its spec

  completed successfully:
  Code Quality#17694:Pull request#3135synchronizebyshengliangxu
  shengliangx/hfep-9a2-ep-spec-applier

  Oct 31, 20265m 19s
  More actions for workflow run 17694

• [OMNIML-6076] [2/3] Prepare a model's config for expert parallelism from its spec

  completed successfully:
  Code Quality#17693:Pull request#3135synchronizebyshengliangxu
  shengliangx/hfep-9a2-ep-spec-applier

  Oct 31, 20264m 25s
  More actions for workflow run 17693

• Code Quality

  completed successfully:
  Code Quality#17692:Scheduled
  main

  Oct 31, 20265m 0s
  More actions for workflow run 17692

• [OMNIML-6073] Quantize Nemotron-H's experts under grouped_mm, with their wrapper in its model folder

  completed successfully:
  Code Quality#17691:Pull request#3067synchronizebyshengliangxu
  shengliangx/hfep-8b-nemotron-h-experts

  Oct 30, 20264m 51s
  More actions for workflow run 17691

• [OMNIML-6071] Calibrate a distributed model on per-rank data

  completed successfully:
  Code Quality#17690:Pull request#2740synchronizebyshengliangxu
  shengliangx/hfep-5-distributed-calibration

  Oct 30, 20265m 44s
  More actions for workflow run 17690

• [OMNIML-6054] [1/2] Move fsdp2_wrap out of distributed.py into fsdp2_utils

  completed successfully:
  Code Quality#17689:Pull request#3140synchronizebyshengliangxu
  shengliangx/hfep-13a-move-fsdp2-wrap

  Oct 30, 20266m 6s
  More actions for workflow run 17689

• [OMNIML-6076] [2/3] Prepare a model's config for expert parallelism from its spec

  completed successfully:
  Code Quality#17688:Pull request#3135synchronizebyshengliangxu
  shengliangx/hfep-9a2-ep-spec-applier

  Oct 30, 20264m 11s
  More actions for workflow run 17688

• Collect imatrix statistics with one fused float32 reduction

  completed successfully:
  Code Quality#17687:Pull request#3044synchronizebycjluo-nv
  chenjiel/imatrix-fused-reduction

  Oct 30, 20265m 0s
  More actions for workflow run 17687

• [OMNIML-6071] Calibrate a distributed model on per-rank data

  completed successfully:
  Code Quality#17686:Pull request#2740synchronizebyshengliangxu
  shengliangx/hfep-5-distributed-calibration

  Oct 30, 20265m 7s
  More actions for workflow run 17686

• [OMNIML-6073] Quantize Nemotron-H's experts under grouped_mm, with their wrapper in its model folder

  completed successfully:
  Code Quality#17685:Pull request#3067synchronizebyshengliangxu
  shengliangx/hfep-8b-nemotron-h-experts

  Oct 30, 20265m 25s
  More actions for workflow run 17685

• Run fused-MoE expert GPTQ updates on all of this process's GPUs

  completed successfully:
  Code Quality#17684:Pull request#3002synchronizebycjluo-nv
  chenjiel/gptq-parallel-expert-updates

  Oct 30, 20265m 58s
  More actions for workflow run 17684

• fix(tools/mcp): stop launcher subprocesses inheriting server stdin

  requires action with the application:
  Code Quality#17683:Pull request#2585synchronizebysomtri
  somtri:fix/2551-launcher-stdin-devnull

  Oct 30, 2026Action required
  More actions for workflow run 17683

• [OMNIML-6073] [2/2] Calibrate fused experts on grouped_mm and torch_a2a instead of forcing eager

  completed successfully:
  Code Quality#17682:Pull request#3074synchronizebyshengliangxu
  shengliangx/hfep-8a-grouped-mm-calibration

  Oct 30, 20264m 35s
  More actions for workflow run 17682

• [OMNIML-6076] [2/3] Prepare a model's config for expert parallelism from its spec

  completed successfully:
  Code Quality#17681:Pull request#3135synchronizebyshengliangxu
  shengliangx/hfep-9a2-ep-spec-applier

  Oct 30, 20265m 42s
  More actions for workflow run 17681