Issue Labeler

Workflow runs for Issue Labeler

• Third-order DPM++ multistep coefficients fail a quadratic forcing test

  completed successfully:
  Issue Labeler#391:Issue#15056openedbyRobertboy18
  Oct 15, 202617s
  More actions for workflow run 391

• [LTX-2] Modular pipelines: let callers skip the audio decode for video-only outputs

  completed successfully:
  Issue Labeler#390:Issue#15044openedbychristopher5106
  Oct 15, 202616s
  More actions for workflow run 390

• Warn when add_noise runs before set_timesteps

  completed successfully:
  Issue Labeler#389:Issue#15043openedbyTejasS1233
  Oct 15, 202619s
  More actions for workflow run 389

• [Wan2.2 VAE, Qwen-Image-2.1 VAE] Redundant x.clone() in residual up/down blocks keeps an extra activation alive

  completed successfully:
  Issue Labeler#388:Issue#15040openedbyluca-888
  Oct 15, 202618s
  More actions for workflow run 388

• LTX2 HDR reference preprocessing applies an unintended stride-aligned resize

  completed successfully:
  Issue Labeler#387:Issue#15037openedbyShenrui-Ma
  Oct 14, 202624s
  More actions for workflow run 387

• fix(ci): harden workflow design flagged on #14789

  failed:
  Issue Labeler#386:Commit154faf9pushedbyhf-security-analysis[bot]Bot
  security/workflow-hardening/pr-14789-design

  Oct 14, 2026Failure
  More actions for workflow run 386

• Qwen Image 2.1 with torch.compile triggers a recompile on each run with varying image size

  completed successfully:
  Issue Labeler#385:Issue#15027openedbygeekuillaume
  Oct 14, 202616s
  More actions for workflow run 385

• SDXL invisible watermark: apply_watermark encodes float input, and about a third of outputs are not detected by Stability's detect.py

  completed successfully:
  Issue Labeler#384:Issue#15025openedbyCyberNativeAI
  Oct 14, 202619s
  More actions for workflow run 384

• Modular pipelines: warn when a denoiser_input_fields entry is not passed to the transformer

  completed successfully:
  Issue Labeler#383:Issue#15022openedbyyiyixuxu
  Oct 13, 202619s
  More actions for workflow run 383

• add_noise() before set_timesteps() uses the wrong noise level (DPMSolverMultistep, DPMSolverSinglestep, UniPC, DEIS; also SanaPipeline)

  completed successfully:
  Issue Labeler#382:Issue#15021openedbyFilipLangiewicz
  Oct 13, 202618s
  More actions for workflow run 382

• [LoRA] Text-encoder LoRA loading assumes CLIP module names and fails for T5 / SmolLM3 / ChatGLM text encoders

  completed successfully:
  Issue Labeler#381:Issue#15020openedbySergasgr
  Oct 12, 202616s
  More actions for workflow run 381

• offload_state_dict=True leaves CPU model weights on meta after loading

  completed successfully:
  Issue Labeler#380:Issue#15018openedbyShenrui-Ma
  Oct 12, 202620s
  More actions for workflow run 380

• gfx1201 (RDNA4) SageAttention attention backend: aiter package or Hub kernel?

  completed successfully:
  Issue Labeler#379:Issue#15014openedbyShan2L
  Oct 12, 202616s
  More actions for workflow run 379

• VaeImageProcessor.denormalize() raises AttributeError for NumPy inputs

  completed successfully:
  Issue Labeler#378:Issue#15011openedbySrijanRoy123-github
  Oct 12, 202616s
  More actions for workflow run 378

• [modular] SequentialPipelineBlocks gets required wrong in both inputs and required_inputs

  completed successfully:
  Issue Labeler#377:Issue#15009openedbyyiyixuxu
  Oct 11, 202616s
  More actions for workflow run 377

• [Z-Image] joint_attention_kwargs={"scale": ...} is silently ignored when loading LoRA weights

  completed successfully:
  Issue Labeler#376:Issue#15007openedbySergasgr
  Oct 11, 202616s
  More actions for workflow run 376

• WanVideoToVideoPipeline: "No cache context is set" when caching is enabled

  completed successfully:
  Issue Labeler#375:Issue#15006openedbyfusheng-ji
  Oct 11, 202617s
  More actions for workflow run 375

• Add TripoSplat

  completed successfully:
  Issue Labeler#374:Issue#15005openedbyhlky
  Oct 11, 202618s
  More actions for workflow run 374

• KDPM2DiscreteScheduler produces NaN on final step due to zero-sigma interpolation

  completed successfully:
  Issue Labeler#373:Issue#15002openedbyriparanta
  Oct 10, 202615s
  More actions for workflow run 373

• load_video() leaks temporary files when remote video decoding fails

  completed successfully:
  Issue Labeler#372:Issue#14994openedbySrijanRoy123-github
  Oct 9, 202616s
  More actions for workflow run 372

• WanCausalConv3d copies every activation with F.pad. Let the conv do the spatial padding.

  completed successfully:
  Issue Labeler#371:Issue#14993openedbywill-rice
  Oct 8, 202623s
  More actions for workflow run 371

• Add an optional Transformer Engine FP8 training example for FLUX

  completed successfully:
  Issue Labeler#370:Issue#14992openedbysbhavani
  Oct 8, 202627s
  More actions for workflow run 370

• MiniMax-H3: _decode_video_file turns ±90° display-rotated videos upside down

  completed successfully:
  Issue Labeler#369:Issue#14985openedbyhsbcelebi
  Oct 7, 202615s
  More actions for workflow run 369

• LTX-2.5 SDR-To-HDR IC-LoRA: how should it be structured?

  failed:
  Issue Labeler#368:Issue#14981openedbychristopher5106
  Oct 7, 202632s
  More actions for workflow run 368

• LTX-2: decouple the RoPE time base from the playback frame rate (Slow-Motion-Control LoRA)

  failed:
  Issue Labeler#367:Issue#14980openedbychristopher5106
  Oct 7, 20261m 28s
  More actions for workflow run 367