All workflows

Workflow runs

• fix(datasets): build the DP all_reduce tensor on the mesh device instead of assuming CUDA

  skipped:
  Claude Code#3991:Issue comment#2127 (comment)createdbyli-lizhe
  Oct 31, 20261s
  More actions for workflow run 3991

• fix(datasets): build the DP all_reduce tensor on the mesh device instead of assuming CUDA

  requires action with the application:
  lint#9430:Pull request#2127reopenedbyli-lizhe
  li-lizhe:fix/dp-allreduce-device

  Oct 31, 2026Action required
  More actions for workflow run 9430

• fix(datasets): build the DP all_reduce tensor on the mesh device instead of assuming CUDA

  requires action with the application:
  unit_test#3437:Pull request#2127reopenedbyli-lizhe
  li-lizhe:fix/dp-allreduce-device

  Oct 31, 2026Action required
  More actions for workflow run 3437

• Claude Code

  skipped:
  Claude Code#3990:createdbyli-lizhe
  Oct 31, 20261s
  More actions for workflow run 3990

• [Fix] fla_npu >= 26.10 compatibility for the GDN AscendC stack (chunk loader + causal conv backends)

  requires action with the application:
  lint#9429:Pull request#2240synchronizebylongboat2010
  longboat2010:fix/fla-npu-2610-compat

  Oct 31, 2026Action required
  More actions for workflow run 9429

• [Fix] fla_npu >= 26.10 compatibility for the GDN AscendC stack (chunk loader + causal conv backends)

  requires action with the application:
  unit_test#3436:Pull request#2240synchronizebylongboat2010
  longboat2010:fix/fla-npu-2610-compat

  Oct 31, 2026Action required
  More actions for workflow run 3436

• fix(datasets): build the DP all_reduce tensor on the mesh device instead of assuming CUDA

  failed:
  lint#9428:Pull request#2127synchronizebyli-lizhe
  li-lizhe:fix/dp-allreduce-device

  Oct 31, 20261h 27m 16s
  More actions for workflow run 9428

• fix(datasets): build the DP all_reduce tensor on the mesh device instead of assuming CUDA

  failed:
  unit_test#3435:Pull request#2127synchronizebyli-lizhe
  li-lizhe:fix/dp-allreduce-device

  Oct 31, 20261h 27m 15s
  More actions for workflow run 3435

• [Fix] fla_npu >= 26.10 compatibility for the GDN AscendC stack (chunk loader + causal conv backends)

  requires action with the application:
  unit_test#3434:Pull request#2240synchronizebylongboat2010
  longboat2010:fix/fla-npu-2610-compat

  Oct 31, 2026Action required
  More actions for workflow run 3434

• [Fix] fla_npu >= 26.10 compatibility for the GDN AscendC stack (chunk loader + causal conv backends)

  requires action with the application:
  lint#9427:Pull request#2240synchronizebylongboat2010
  longboat2010:fix/fla-npu-2610-compat

  Oct 31, 2026Action required
  More actions for workflow run 9427

• ete_test_npu-all

  cancelled:
  ete_test_npu#262:Scheduled
  main

  Oct 30, 20266h 0m 53s
  More actions for workflow run 262

• weekly-ete-test-all

  failed:
  weekly ete test#34:Scheduled
  main

  Oct 30, 202619h 17m 51s
  More actions for workflow run 34

• weekly-ete-test-all

  failed:
  weekly ete test#33:Manually runbyzhulinJulia24
  update_lmdeploy

  Oct 30, 202617h 38m 58s
  More actions for workflow run 33

• [Refactor] Reuse single-media pack tensors without torch.cat copies

  cancelled:
  unit_test#3433:Pull request#2239synchronizebyShilohYu
  ShilohYu:feat/glm53-media-cat

  Oct 30, 20266h 39m 24s
  More actions for workflow run 3433

• [Refactor] Reuse single-media pack tensors without torch.cat copies

  completed successfully:
  lint#9426:Pull request#2239synchronizebyShilohYu
  ShilohYu:feat/glm53-media-cat

  Oct 30, 20262m 5s
  More actions for workflow run 9426

• [Feature] Add opt-in SP leader-only full-batch loading

  skipped:
  Claude Code#3989:Pull request#2223submittedbyShilohYu
  Oct 30, 20269s
  More actions for workflow run 3989

• [Feature] Add opt-in SP leader-only full-batch loading

  skipped:
  Claude Code#3988:Pull request#2223createdbyShilohYu
  Oct 30, 202610s
  More actions for workflow run 3988

• [Feature] Add opt-in SP leader-only full-batch loading

  skipped:
  Claude Code#3987:Pull request#2223createdbyShilohYu
  Oct 30, 20266s
  More actions for workflow run 3987

• [Feature] Add opt-in SP leader-only full-batch loading

  skipped:
  Claude Code#3986:Pull request#2223submittedbyShilohYu
  Oct 30, 20261s
  More actions for workflow run 3986

• [Docs] Harden align-automodel-xtuner-text skill from field runs and cross-model reviews

  cancelled:
  unit_test#3432:Pull request#2232synchronizebyDenny991
  Denny991:fix/text-alignment-field-hardening

  Oct 30, 20269h 1m 47s
  More actions for workflow run 3432

• [Docs] Harden align-automodel-xtuner-text skill from field runs and cross-model reviews

  completed successfully:
  lint#9425:Pull request#2232synchronizebyDenny991
  Denny991:fix/text-alignment-field-hardening

  Oct 30, 20262m 18s
  More actions for workflow run 9425

• [Docs] Harden align-automodel-xtuner-text skill from field runs and cross-model reviews

  cancelled:
  unit_test#3431:Pull request#2232synchronizebyDenny991
  Denny991:fix/text-alignment-field-hardening

  Oct 30, 20263m 6s
  More actions for workflow run 3431

• [Docs] Harden align-automodel-xtuner-text skill from field runs and cross-model reviews

  completed successfully:
  lint#9424:Pull request#2232synchronizebyDenny991
  Denny991:fix/text-alignment-field-hardening

  Oct 30, 20262m 26s
  More actions for workflow run 9424

• publish-docker-inner

  completed successfully:
  publish-docker-inner#263:Manually runbyYifanHe-ailab
  heyifan/update_env_1019

  Oct 30, 20261h 45m 33s
  More actions for workflow run 263

• [feat] Support value head and critic loss

  skipped:
  Claude Code#3985:Pull request#2188createdbyPengchengShi00
  Oct 30, 20261s
  More actions for workflow run 3985