Issues

• Automatically reusing components

  idea-contribution-welcome
  huggingface/diffusers#15248 · Filter by author yiyixuxu opened on Nov 11, 2026

• [MiniMax-H3, Pika H3 Flex] ref2va fails on a reference soundtrack that is not at 32 kHz without torchaudio

  huggingface/diffusers#15245 · Filter by author christopher5106 opened on Nov 11, 2026

• LLaDA2: EOS early stopping runs every remaining block to the step limit in batched generation

  bug
  huggingface/diffusers#15242 · Filter by author yyu-260 opened on Nov 11, 2026

• Stable Audio 3 attention repeats RoPE cos / sin and q / k norm work on every call

  huggingface/diffusers#15239 · Filter by author chsu-amd opened on Nov 11, 2026

• Docs: 17 dead external links (verified ×3) — upstream content removed, needs maintainer decisions

  huggingface/diffusers#15233 · Filter by author Hisernberg opened on Nov 10, 2026

• How should ModularPipeline handle a component that failed to load?

  idea-contribution-welcome
  huggingface/diffusers#15227 · Filter by author yiyixuxu opened on Nov 9, 2026
  3

• Wan transformer block makes float32 copies of hidden_states in its residual adds

  huggingface/diffusers#15224 · Filter by author klin2024 opened on Nov 9, 2026

• FP32LayerNorm makes two full-size float32 copies on every call

  huggingface/diffusers#15221 · Filter by author klin2024 opened on Nov 9, 2026

• RoPE helpers make float32 copies of q / k on every call

  huggingface/diffusers#15220 · Filter by author klin2024 opened on Nov 9, 2026

• _tp_plan / _cp_plan: allow entries for modules that are not always present

  idea-contribution-welcomeperformance
  huggingface/diffusers#15218 · Filter by author yiyixuxu opened on Nov 9, 2026 · Release 0.44.0
  8

  

• unload_textual_inversion(token) leaves the extra tokens of a multi-vector embedding behind, so it can't be loaded again

  huggingface/diffusers#15211 · Filter by author sohailshk opened on Nov 8, 2026
  3

• set_ip_adapter_scale(float) raises "Cannot assign 1 scale_configs to 2 IP-Adapter" when several IP-Adapters are loaded

  huggingface/diffusers#15210 · Filter by author sohailshk opened on Nov 8, 2026

• [modular] Per-workflow input signature in the docstring of blocks with _workflow_map

  idea-contribution-welcome
  huggingface/diffusers#15208 · Filter by author yiyixuxu opened on Nov 8, 2026
  3

• UniPCMultistepScheduler.set_timesteps(sigmas=...) raises TypeError for a list when use_flow_sigmas=True

  bug
  huggingface/diffusers#15207 · Filter by author charchit7 opened on Nov 7, 2026

• device_map="balanced" drops all but the last CPU-overflow component, which then loads without offload hooks and breaks inference

  bug
  huggingface/diffusers#15204 · Filter by author charchit7 opened on Nov 7, 2026
  3

• FLEX attention silently ignores is_causal=True when no mask is supplied

  huggingface/diffusers#15195 · Filter by author davrens101 opened on Nov 6, 2026

• HookRegistry.remove_hook drops the next hook's new_forward (breaks FirstBlockCache with group offloading)

  huggingface/diffusers#15193 · Filter by author theontho opened on Nov 6, 2026

• Per-token FlowMatch Euler steps use the opposite velocity sign

  huggingface/diffusers#15185 · Filter by author davrens101 opened on Nov 5, 2026
  1

• flash_varlen raises when requesting log-sum-exp

  huggingface/diffusers#15182 · Filter by author 0z5a opened on Nov 4, 2026

• flash_4_hub does not forward return_lse to its kernel

  huggingface/diffusers#15180 · Filter by author 0z5a opened on Nov 4, 2026

• img2img: outputs paired with the wrong init image

  bug
  huggingface/diffusers#15176 · Filter by author Arnav66692 opened on Nov 3, 2026
  1

• Cosmos 3 docs: FP8 snippet fails at the first W8A8 step without importing ModelOpt's FP8 GEMM backend

  huggingface/diffusers#15170 · Filter by author piotrm-nvidia opened on Nov 2, 2026
  1

• Cosmos 3 FP8: wrapping pipe.transformer (e.g. torch.compile) silently disables the checkpoint's W8A16 step policy

  huggingface/diffusers#15169 · Filter by author piotrm-nvidia opened on Nov 2, 2026

• UniPC and DPM-Solver with use_flow_sigmas condition the model on integer timesteps while integrating with float sigmas

  huggingface/diffusers#15167 · Filter by author LiRunGuo opened on Nov 2, 2026
  2

• [training scripts] --logit_mean biases timesteps in the opposite direction to the SD3 paper (positive m oversamples low noise)

  huggingface/diffusers#15166 · Filter by author Andreas-Hum opened on Nov 1, 2026
  3