pypi-diffusers: Autospec creation for update from version 0.32.2 to version 0.33.0

39th president of the United States, probably (1):
      Fix for multi-GPU WAN inference (#10997)

Abhipsha Das (1):
      [Model Card] standardize advanced diffusion training sdxl lora (#7615)

Ahmed Belgacem (1):
      Fix redundant prev_output_channel assignment in UNet2DModel (#10945)

Alan Ponnachan (1):
      Add torch_xla support to pipeline_aura_flow.py (#10365)

Alexey Zolotenkov (1):
      Fix incorrect seed initialization when args.seed is 0 (#10964)

Ameer Azam (1):
      Regarding the RunwayML path for V1.5 did change to stable-diffusion-v1-5/[stable-diffusion-v1-5/ stable-diffusion-inpainting] (#10476)

Andreas Jörg (1):
      [examples/controlnet/train_controlnet_sd3.py] Fixes #11050 - Cast prompt_embeds and pooled_prompt_embeds to weight_dtype to prevent dtype mismatch (#11051)

Anton Obukhov (1):
      Marigold Update: v1-1 models, Intrinsic Image Decomposition pipeline, documentation (#10884)

Aryan (35):
      Fix TorchAO related bugs; revert device_map changes (#10371)
      Update Flux docstrings (#10423)
      Update variable names correctly in docs (#10435)
      Fix hunyuan video attention mask dim (#10454)
      Fix style (#10478)
      [LoRA] Support original format loras for HunyuanVideo (#10376)
      Add `_no_split_modules` to some models (#10308)
      Fix Latte output_type (#10558)
      Test sequential cpu offload for torchao quantization (#10506)
      Fix offload tests for CogVideoX and CogView3 (#10547)
      [core] Layerwise Upcasting (#10347)
      Improve TorchAO error message (#10627)
      [core] Pyramid Attention Broadcast (#9562)
      Refactor gradient checkpointing (#10611)
      Refactor OmniGen (#10771)
      Disable PEFT input autocast when using fp8 layerwise casting (#10685)
      Update FlowMatch docstrings to mention correct output classes (#10788)
      Refactor CogVideoX transformer forward (#10789)
      Module Group Offloading (#10503)
      Remove print statements (#10836)
      Some consistency-related fixes for HunyuanVideo (#10835)
      SkyReels Hunyuan T2V & I2V (#10837)
      [docs] Add CogVideoX Schedulers (#10885)
      [refactor] SD3 docs & remove additional code (#10882)
      [refactor] Remove additional Flux code (#10881)
      [LoRA] Support Wan (#10943)
      Hunyuan I2V (#10983)
      [LoRA] CogView4 (#10981)
      LTX 0.9.5 (#10968)
      Group offloading improvements (#11094)
      Fix Group offloading behaviour when using streams (#11097)
      [core] FasterCache (#10163)
      New HunyuanVideo-I2V (#11066)
      Improve information about group offloading and layerwise casting (#11101)
      Raise warning and round down if Wan num_frames is not 4k + 1 (#11167)

AstraliteHeart (2):
      Add AuraFlow GGUF support (#10463)
      Add missing `isinstance` for arg checks in GGUFParameter (#10834)

Bagheera (1):
      fix for #7365, prevent pipelines from overriding provided prompt embeds (#7926)

Basile Lewandowski (1):
      Change KolorsPipeline LoRA Loader to StableDiffusion (#11198)

Benjamin Bossan (1):
      [LoRA] Implement hot-swapping of LoRA (#9453)

Bhavay Malhotra (1):
      [train_controlnet.py] Fix the LR schedulers when num_train_epochs is passed in a distributed training env (#8461)

Bruno Magalhaes (1):
      remove unnecessary call to `F.pad` (#10620)

Bubbliiiing (1):
      Add EasyAnimateV5.1 text-to-video, image-to-video, control-to-video generation model  (#10626)

C (3):
      [Docs] Add documentation about using ParaAttention to optimize FLUX and HunyuanVideo (#10544)
      Fix Graph Breaks When Compiling CogView4 (#10959)
      Fix Wan I2V Quality (#11087)

Cheng Jin (1):
      Resolve stride mismatch in UNet's ResNet to support Torch DDP (#11098)

CyberVy (7):
      Fix Callback Tensor Inputs of the SDXL Controlnet Inpaint and Img2img Pipelines are missing "controlnet_image". (#10880)
      Fix Callback Tensor Inputs of the SD Controlnet Pipelines are missing some elements. (#10907)
      Improve load_ip_adapter RAM Usage (#10948)
      Fix the missing parentheses when calling is_torchao_available in quantization_config.py. (#10961)
      Fix Flux Controlnet Pipeline _callback_tensor_inputs Missing Some Elements (#10974)
      Fix missing **kwargs in lora_pipeline.py (#11011)
      fix _callback_tensor_inputs of sd controlnet inpaint pipeline missing some elements (#11073)

Célina (1):
      use style bot GH Action from `huggingface_hub` (#10970)

Daniel Hipke (1):
      Add a `disable_mmap` option to the `from_single_file` loader to improve load performance on network mounts (#10305)

Daniel Regado (9):
      IP-Adapter support for `StableDiffusion3ControlNetPipeline` (#10363)
      Added IP-Adapter for `StableDiffusion3ControlNetInpaintingPipeline` (#10561)
      IP-Adapter for `StableDiffusion3InpaintPipeline` (#10581)
      IP-Adapter for `StableDiffusion3Img2ImgPipeline` (#10589)
      [Docs] Update SD3 ip_adapter model_id to diffusers checkpoint (#10597)
      `MultiControlNetUnionModel` on SDXL (#10747)
      SD3 IP-Adapter runtime checkpoint conversion (#10718)
      Comprehensive type checking for `from_pretrained` kwargs (#10758)
      Multi IP-Adapter for Flux pipelines (#10867)

Dev Rajput (1):
      Add correct number of channels when resuming from checkpoint for Flux Control LoRa training (#10422)

Dhruv Nair (28):
      [CI] Add minimal testing for legacy Torch versions (#10479)
      [CI] Torch Min Version Test Fix (#10491)
      [Single File] Fix loading Flux Dev finetunes with Comfy Prefix (#10545)
      [CI] Update HF Token in Fast GPU Tests  (#10568)
      [CI] Update HF Token on Fast GPU Model Tests (#10570)
      [CI] Update HF_TOKEN in all workflows (#10613)
      [CI] Fix Truffle Hog failure (#10769)
      [Single File] Add Single File support for Lumina Image 2.0 Transformer (#10781)
      [CI] Fix incorrectly named test module for Hunyuan DiT (#10854)
      [CI] Update always test Pipelines list in Pipeline fetcher (#10856)
      [Docs] Fix toctree sorting  (#10894)
      [CI] Improvements to conditional GPU PR tests (#10859)
      [CI] Fix Fast GPU tests on PR (#10912)
      [CI] Fix for failing IP Adapter test in Fast GPU PR tests (#10915)
      [CI] Update Stylebot Permissions (#10931)
      [Single File] Add user agent to SF download requests.  (#10979)
      [Single File] Add single file support for Wan T2V/I2V (#10991)
      Fix for fetching variants only (#10646)
      [Quantization] Add Quanto backend (#10756)
      [Quantization] Allow loading TorchAO serialized Tensor objects with torch>=2.6  (#11018)
      [Refactor] Clean up import utils boilerplate (#11026)
      Provide option to reduce CPU RAM usage in Group Offload (#11106)
      [Quantization] dtype fix for GGUF + fix BnB tests (#11159)
      [Docs] Update Wan Docs with memory optimizations (#11089)
      [WIP] Add Wan Video2Video (#11053)
      Add CacheMixin to Wan and LTX Transformers (#11187)
      Fix Single File loading for LTX VAE  (#11200)
      Update Ruff to latest Version (#10919)

Dimitri Barbot (2):
      Fix pipeline dtype unexpected change when using SDXL reference community pipelines in float16 mode (#10670)
      Fix deterministic issue when getting pipeline dtype and device (#10696)

Doug J (1):
      Update train_text_to_image_sdxl.py (#8830)

Edna (1):
      Add Wan with STG as a community pipeline (#11184)

Eliseu Silva (7):
      Make passing the IP Adapter mask to the attention mechanism optional (#10346)
      feat: new community mixture_tiling_sdxl pipeline for SDXL (#10759)
      fix: [Community pipeline] Fix flattened elements on image  (#10774)
      feat: add Mixture-of-Diffusers ControlNet Tile upscaler Pipeline for SDXL (#10951)
      fix: mixture tiling sdxl pipeline - adjust gerating time_ids & embeddings  (#11012)
      fix: for checking mandatory and optional pipeline components (#11189)
      feat: [Community Pipeline] - FaithDiff Stable Diffusion XL Pipeline (#11188)

Fanli Lin (8):
      [tests] fix `AssertionError: Torch not compiled with CUDA enabled` (#10356)
      [tests] make tests device-agnostic (part 3)  (#10437)
      make tensors contiguous before passing to safetensors (#10761)
      [tests] make tests device-agnostic (part 4) (#10508)
      [tests] enable bnb tests on xpu (#11001)
      [tests] make cuda only tests device-agnostic  (#11058)
      [tests] no hard-coded cuda  (#11186)
      [tests] HunyuanDiTControlNetPipeline inference precision issue on XPU (#11197)

G.O.D (1):
      fix bug for ascend npu (#10429)

Giuseppe Catalano (1):
      Revert RePaint scheduler 'fix' (#10644)

Hanch Han (1):
      [fix] refer use_framewise_encoding on AutoencoderKLHunyuanVideo._encode (#10600)

Haoyun Qin (1):
      fix: support transformer models' `generation_config` in pipeline (#10779)

Ikpreet S Babra (1):
      Fixed grammar in "write_own_pipeline" readme (#10706)

Ilya Drobyshevskiy (1):
      fix flux controlnet bug (#11152)

Inigo Goiri (1):
      Add support to pass image embeddings to the WAN I2V pipeline. (#11175)

Ishan Modi (1):
      [Single File] Add single file loading for SANA Transformer (#10947)

Jacob Helwig (1):
      Add sigmoid scheduler in `scheduling_ddpm.py` docs (#10648)

Juan Acevedo (3):
      implementing flux on TPUs with ptxla (#10515)
      reverts accidental change that removes attn_mask in attn. Improves fl… (#11065)
      update readme instructions. (#11096)

Jun Yeop Na (2):
      [train_dreambooth_lora.py] Fix the LR Schedulers when `num_train_epochs` is passed in a distributed training env (#10973)
      [doc] Fix Korean Controlnet Train doc (#11141)

Junsong Chen (8):
      [Sana] 1k PE bug fixed (#10431)
      [Sana][bug fix]change clean_caption from True to False. (#10481)
      [Sana 4K] (#10493)
      [Sana] add Sana to auto-text2image-pipeline; (#10538)
      [Sana-4K] (#10537)
      [fix bug] PixArt inference_steps=1 (#11079)
      [fix SANA-Sprint] (#11142)
      add a timestep scale for sana-sprint teacher model (#11150)

Junyu Chen (2):
      [DC-AE] support tiling for DC-AE (#10510)
      [DC-AE, SANA] fix SanaMultiscaleLinearAttention apply_quadratic_attention bf16 (#10595)

Kenneth Gerald Hamilton (1):
      Fixed requests.get function call by adding timeout parameter. (#11156)

Kinam Kim (1):
      Add STG to community pipelines (#10960)

Le Zhuo (1):
      Add support for lumina2 (#10642)

Leo Jiang (4):
      [Sana 4K] Add vae tiling option to avoid OOM (#10583)
      NPU adaption for RMSNorm (#10534)
      NPU Adaption for Sanna (#10409)
      [bugfix] NPU Adaption for Sana (#10724)

Linoy Tsaban (4):
      small readme changes for advanced training examples (#10473)
      [flux lora training] fix t5 training bug (#10845)
      [Wan LoRAs] make T2V LoRAs compatible with Wan I2V (#11107)
      [Flux LoRA] fix issues in flux lora scripts (#11111)

LittleNyima (1):
      Add CogVideoX DDIM Inversion to Community Pipelines (#10956)

Lucain (1):
      Remove cache migration script (#10619)

Luchao Qi (1):
      [Typo] Update md files (#10404)

Marc Sun (5):
      Fix compatibility with pipeline when loading model with device_map on single gpu (#10390)
      [FEAT] DDUF format (#10037)
      [FEAT] Model loading refactor (#10604)
      store activation cls instead of function (#10832)
      remove format check for safetensors file (#10864)

Mark (1):
      [Docs] Fix environment variables in `installation.md` (#11179)

Marlon May (1):
      Add community pipeline for semantic guidance for FLUX (#10610)

Mathias Parger (1):
      speedup hunyuan encoder causal mask generation (#10764)

Max Podkorytov (1):
      Fix enable memory efficient attention on ROCm (#10564)

Mikko Tukiainen (1):
      Add missing MochiEncoder3D.gradient_checkpointing attribute (#11146)

Muyang Li (2):
      Use `randn_tensor` to replace `torch.randn` (#10535)
      Remove the FP32 Wrapper when evaluating (#10617)

Nicolas (1):
      Fix train_text_to_image.py --help (#10711)

Omar Awile (1):
      Fix documentation for FluxPipeline (#10563)

Parag Ekbote (7):
      Notebooks for Community Scripts-5 (#10499)
      Fix Documentation about Image-to-Image Pipeline (#10704)
      Notebooks for Community Scripts-6 (#10713)
      Extend Support for callback_on_step_end for AuraFlow and LuminaText2Img Pipelines (#10746)
      Notebooks for Community Scripts-7 (#10846)
      Add Example of IPAdapterScaleCutoffCallback to Docs (#10934)
      Notebooks for Community Scripts-8 (#11128)

Pierre Chapuis (1):
      fix default values of Flux guidance_scale in docstrings (#10982)

Rahul Raman (1):
      Refactor instructpix2pix lora to support peft (#10205)

Raul Ciotescu (1):
      width and height are mixed-up (#10629)

SahilCarterr (8):
      [Add] torch_xla support to pipeline_sana.py (#10364)
      [Fix] Broken links in hunyuan docs (#10402)
      [Fix] broken links in docs (#10434)
      [FIX] check_inputs function in Auraflow Pipeline (#10678)
      [Fix] Type Hint in from_pretrained() to Ensure Correct Type Inference (#10714)
      [FIX] check_inputs function in lumina2 (#10784)
      [Fix] Docs overview.md (#10858)
      [Fix] fp16 unscaling in train_dreambooth_lora_sdxl (#10889)

Sayak Paul (75):
      [chore] post release 0.32.0 (#10361)
      fix test pypi installation in the release workflow (#10360)
      [training] fix: registration of out_channels in the control flux scripts. (#10367)
      [LoRA] feat: support `unload_lora_weights()` for Flux Control. (#10206)
      [training] add ds support to lora sd3. (#10378)
      [LTX-Video] fix attribute adjustment for ltx. (#10426)
      [Tests] add slow and nightly markers to sd3 lora integation. (#10458)
      [LoRA] fix: lora unloading when using expanded Flux LoRAs. (#10397)
      [Training] QoL improvements in the Flux Control training scripts (#10461)
      [chore] remove prints from tests. (#10505)
      [LoRA] clean up `load_lora_into_text_encoder()` and `fuse_lora()` copied from (#10495)
      [LoRA] allow big CUDA tests to run properly for LoRA (and others) (#9845)
      [CI] Match remaining assertions from big runner (#10521)
      [Tests] skip tests properly with `unittest.skip()` (#10527)
      [Docs] Add negative prompt docs to FluxPipeline (#10531)
      [Flux] Improve true cfg condition (#10539)
      [LoRA] improve failure handling for peft. (#10551)
      [Docs] Update hunyuan_video.md to rectify the checkpoint id (#10524)
      [LoRA] feat: support loading loras into 4bit quantized Flux models. (#10578)
      [Tests] add: test to check 8bit bnb quantized models work with lora loading. (#10576)
      [Chore] fix vae annotation in mochi pipeline (#10585)
      [training] set rest of the blocks with `requires_grad` False. (#10607)
      [chore] change licensing to 2025 from 2024. (#10615)
      [Tests] modify the test slices for the failing flax test (#10630)
      [docs] fix image path in para attention docs (#10632)
      [chore] add a script to extract loras from full fine-tuned models (#10631)
      [Tests] conditionally check `fp8_e4m3_bf16_max_memory < fp8_e4m3_fp32_max_memory`  (#10669)
      [tests] update llamatokenizer in hunyuanvideo tests (#10681)
      [bitsandbytes] Simplify bnb int8 dequant (#10401)
      [LoRA] fix peft state dict parsing (#10532)
      [Tests] Test layerwise casting with training (#10765)
      [chore] update notes generation spaces (#10592)
      [LoRA] improve lora support for flux. (#10810)
      [docs] add missing entries to the lora docs. (#10819)
      [LoRA] make `set_adapters()` robust on silent failures. (#9618)
      [misc] feat: introduce a style bot. (#10274)
      [tests] use proper gemma class and config in lumina2 tests. (#10828)
      [LoRA] add LoRA support to Lumina2 and fine-tuning script (#10818)
      [Utils] add utilities for checking if certain utilities are properly documented (#7763)
      [tests] test `encode_prompt()` in isolation (#10438)
      [CI] install accelerate transformers from `main` (#10289)
      [CI] run fast gpu tests conditionally on pull requests. (#10310)
      fix: run tests from a pr workflow. (#9696)
      [chore] template for remote vae. (#10849)
      fix remote vae template (#10852)
      [LoRA] restrict certain keys to be checked for peft config update. (#10808)
      [chore] correct qk norm list. (#10876)
      [Tests] fix: lumina2 lora fuse_nan test (#10911)
      [style bot] improve security for the stylebot. (#10908)
      [chore] fix-copies to flux pipelines (#10941)
      [Tests] Remove more encode prompts tests (#10942)
      Update evaluation.md (#10938)
      [LoRA] feat: support non-diffusers lumina2 LoRAs. (#10909)
      [tests] fix tests for save load components (#10977)
      [CI] remove synchornized. (#10980)
      [LoRA] remove full key prefix from peft. (#11004)
      [LoRA] Improve copied from comments in the LoRA loader classes (#10995)
      [LoRA] Improve warning messages when LoRA loading becomes a no-op (#10187)
      [Tests] improve quantization tests by additionally measuring the inference memory savings (#11021)
      [LoRA] support wan i2v loras from the world. (#11025)
      [LoRA] change to warning from info when notifying the users about a LoRA no-op (#11044)
      [Tests] restrict memory tests for quanto for certain schemes. (#11052)
      [LoRA] feat: support non-diffusers wan t2v loras. (#11059)
      [Tests] add requires peft decorator. (#11037)
      [CI] pin transformers version for benchmarking. (#11067)
      make PR GPU tests conditioned on styling. (#11099)
      [CI] uninstall deps properly from pr gpu tests. (#11102)
      [feat] implement `record_stream` when using CUDA streams during group offloading (#11081)
      [bistandbytes] improve replacement warnings for bnb (#11132)
      minor update to sana sprint docs. (#11236)
      [docs] minor updates to dtype map docs. (#11237)
      [LoRA] support more comyui loras for Flux 🚨 (#10985)
      fix: SD3 ControlNet validation so that it runs on a A100. (#11238)
      fix timeout constant (#11252)
      fix consisid imports (#11254)

Shenghai Yuan (1):
      [core] ConsisID (#10140)

Shitao Xiao (1):
      Add OmniGen (#10148)

Steven Liu (9):
      [docs] Quantization tip (#10249)
      [docs] Video generation update (#10272)
      [docs] Fix internal links (#10418)
      [docs] Fix missing parameters in docstrings (#10419)
      [docs] uv installation (#10622)
      [docs] LoRA support (#10844)
      [docs] Update prompt weighting docs (#10843)
      [docs] Flux group offload (#10847)
      [docs] MPS update (#11212)

Suprhimp (1):
      [feat]Add strength in flux_fill pipeline (denoising strength for fluxfill) (#10603)

Teriks (5):
      RFInversionFluxPipeline, small fix for enable_model_cpu_offload & enable_sequential_cpu_offload compatibility (#10480)
      allow passing hf_token to load_textual_inversion (#10546)
      SDXL ControlNet Union pipelines, make control_image argument immutible (#10663)
      support StableDiffusionAdapterPipeline.from_single_file (#10552)
      Fix SD2.X clip single file load projection_dim (#10770)

Thanh Le (2):
      Fix inconsistent random transform in instruct pix2pix (#10698)
      Faster set_adapters (#10777)

Tolga Cangöz (3):
      [`Research Project`] Add AnyText: Multilingual Visual Text Generation And Editing (#8998)
      Update README and example code for AnyText usage (#11028)
      [LTX0.9.5] Refactor `LTXConditionPipeline` for text-only conditioning (#11174)

Vedat Baday (2):
      fix(hunyuan-video): typo in height and width input check (#10684)
      feat(training-utils): support device and dtype params in compute_density_for_timestep_sampling (#10699)

Vinh H. Pham (1):
      Implement framewise encoding/decoding in LTX Video VAE (#10488)

Vladimir Mandic (1):
      flux: make scheduler config params optional (#10384)

Wenhao Sun (1):
      Add pipeline_stable_diffusion_xl_attentive_eraser (#10579)

Yaniv Galron (3):
      removing redundant requires_grad = False (#10628)
      typo fix (#10802)
      making ```formatted_images``` initialization compact (#10801)

Yao Matrix (4):
      map BACKEND_RESET_MAX_MEMORY_ALLOCATED to reset_peak_memory_stats on XPU (#11191)
      enable 1 case on XPU (#11219)
      introduce compute arch specific expectations and fix test_sd3_img2img_inference failure (#11227)
      fix FluxReduxSlowTests::test_flux_redux_inference case failure on XPU (#11245)

YiYi Xu (7):
      make style for huggingface/diffusers#10368 (#10370)
      fix offload gpu tests etc (#10366)
      follow-up refactor on lumina2 (#10776)
      [Alibaba Wan Team] continue on #10921 Wan2.1 (#10922)
      update check_input for cogview4 (#10966)
      remove F.rms_norm for now (#11126)
      add sana-sprint (#11074)

Yih-Dar (1):
      Security fix (#10905)

Yuqian Hong (2):
      create a script to train autoencoderkl (#10605)
      [BUG] Fix Autoencoderkl train script (#11113)

Yuxuan Zhang (5):
      CogView4 (supports different length c and uc) (#10649)
      Update pipeline_cogview4.py (#10944)
      [Docs] CogView4 comment fix (#10957)
      CogView4 Control Block (#10809)
      Modify the implementation of retrieve_timesteps in CogView4-Control. (#11125)

Zehuan Huang (1):
      Support pass kwargs to cogvideox custom attention processor (#10456)

ZhengKai91 (1):
      Fix aclnnRepeatInterleaveIntWithDim error on NPU for get_1d_rotary_pos_embed (#10820)

a120092009 (1):
      [Quantization]  support pass MappingType for TorchAoConfig (#10927)

alex choi (1):
      ensure dtype match between diffused latents and vae weights (#8391)

andreabosisio (1):
      Typo fix in the table number of a referenced paper  (#10528)

baymax591 (1):
      bugfix for npu not support float64 (#10123)

chaowenguo (4):
      add pythor_xla support for render a video (#10443)
      Update rerender_a_video.py fix dtype error (#10451)
      add callable object to convert frame into control_frame to reduce cpu memory usage. (#10501)
      add the xm.mark_step for the first denosing loop (#10530)

co63oc (1):
      Fix pipeline_flux_controlnet.py (#11095)

célina (1):
      Update Style Bot workflow (#11202)

dependabot[bot] (2):
      Bump jinja2 from 3.1.4 to 3.1.5 in /examples/research_projects/realfill (#10377)
      Bump jinja2 from 3.1.5 to 3.1.6 in /examples/research_projects/realfill (#10984)

fancydaddy (1):
      add from_single_file to animatediff  (#10924)

geronimi73 (1):
      AutoModel instead of AutoModelForCausalLM (#10507)

hlky (59):
      Default values in SD3 pipelines when submodules are not loaded (#10393)
      Fix AutoPipeline `from_pipe` where source pipeline is missing target pipeline's optional components (#10400)
      Add torch_xla and from_single_file support to TextToVideoZeroPipeline (#10445)
      `lora_bias` PEFT version check in `unet.load_attn_procs` (#10474)
      LEditsPP - examples, check height/width, add tiling/slicing (#10471)
      Add torch_xla and from_single_file to instruct-pix2pix (#10444)
      Use Pipelines without scheduler (#10439)
      Fix HunyuanVideo produces NaN on PyTorch<2.5 (#10482)
      Use pipelines without vae (#10441)
      Update tokenizers in `pr_test_peft_backend` (#10132)
      Fix tokenizers install from main in LoRA tests (#10494)
      UNet2DModel mid_block_type (#10469)
      PyTorch/XLA support (#10498)
      Use Pipelines without unet (#10440)
      Fix StableDiffusionInstructPix2PixPipelineSingleFileSlowTests (#10557)
      Fix train_dreambooth_lora_sd3_miniature (#10554)
      Fix Nightly AudioLDM2PipelineFastTests (#10556)
      Fix batch > 1 in HunyuanVideo (#10548)
      Move buffers to device (#10523)
      Scheduling fixes on MPS (#10549)
      Add IP-Adapter example to Flux docs (#10633)
      ControlNet Union controlnet_conditioning_scale for multiple control inputs (#10666)
      [training] Convert to ImageFolder script (#10664)
      Add provider_options to OnnxRuntimeModel (#10661)
      Quantized Flux with IP-Adapter (#10728)
      EDMEulerScheduler accept sigmas, add final_sigmas_type (#10734)
      Add `Self` type hint to `ModelMixin`'s `from_pretrained` (#10742)
      Fix `use_lu_lambdas` and `use_karras_sigmas` with `beta_schedule=squaredcos_cap_v2` in `DPMSolverMultistepScheduler` (#10740)
      DiffusionPipeline mixin `to`+FromOriginalModelMixin/FromSingleFileMixin `from_single_file` type hint (#10811)
      `device_map` in `load_model_dict_into_meta` (#10851)
      Fix `torch_dtype` in Kolors text encoder with `transformers` v4.49 (#10816)
      Add SD3 ControlNet to AutoPipeline (#10888)
      Experimental per control type scale for ControlNet Union (#10723)
      Support IPAdapter for more Flux pipelines (#10708)
      Add `remote_decode` to `remote_utils` (#10898)
      Update VAE Decode endpoints (#10939)
      Add VAE Decode endpoint slow test (#10946)
      Fix loading OneTrainer Flux LoRA (#10978)
      Wan VAE move scaling to pipeline (#10998)
      Fix SD3 IPAdapter feature extractor (#11027)
      Use `output_size` in `repeat_interleave` (#11030)
      [hybrid inference 🍯🐝] Add VAE encode (#11017)
      Wan Pipeline scaling fix, type hint warning, multi generator fix (#11007)
      Rename Lumina(2)Text2ImgPipeline -> Lumina(2)Pipeline (#10827)
      Quality options in `export_to_video` (#11090)
      Flux with Remote Encode (#11091)
      Don't override `torch_dtype` and don't use when `quantization_config` is set (#11039)
      WanI2V encode_image (#11164)
      Fix LatteTransformer3DModel dtype mismatch with enable_temporal_attentions (#11139)
      Add `latents_mean` and `latents_std` to `SDXLLongPromptWeightingPipeline` (#11034)
      allow models to run with a user-provided dtype map instead of a single dtype (#10301)
      Revert `save_model` in ModelMixin save_pretrained and use safe_serialization=False in test (#11196)
      [docs] `torch_dtype` map (#11194)
      Fix enable_sequential_cpu_offload in CogView4Pipeline (#11195)
      SchedulerMixin from_pretrained and ConfigMixin Self type annotation (#11192)
      Flux quantized with lora (#10990)
      AudioLDM2 Fixes (#11244)
      AutoModel (#11115)
      [docs] AutoModel (#11250)

jiqing-feng (2):
      Enable dreambooth lora finetune example on other devices (#10602)
      fix autocast (#11190)

kahmed10 (1):
      add onnxruntime-migraphx as part of check for onnxruntime in import_utils.py (#10624)

kakukakujirori (1):
      Bug fix in LTXImageToVideoPipeline.prepare_latents() when latents is already set (#10918)

kentdan3msu (1):
      Set self._hf_peft_config_loaded to True when LoRA is loaded using `load_lora_adapter` in PeftAdapterMixin class (#11155)

lakshay sharma (1):
      Update import_utils.py (#10329)

maxs-kan (1):
      Fix Flux multiple Lora loading bug (#10388)

puhuk (2):
      Update Custom Diffusion Documentation for Multiple Concept Inference to resolve issue #10791 (#10792)
      Fix max_shift value in flux and related functions to 1.15 (issue #10675) (#10807)

sayakpaul (1):
      Release: v0.33.0

sunxunle (1):
      chore: remove redundant words (#10609)

suzukimain (2):
      [Docs] Added `model search` to community_projects.md (#10358)
      [Community] Enhanced `Model Search` (#10417)

victolee0 (1):
      fix check_inputs func in LuminaText2ImgPipeline (#10651)

wonderfan (1):
      chore: fix help messages in advanced diffusion examples (#10923)

xieofxie (1):
      add provider_options in from_pretrained (#10719)

yupeng1111 (1):
      fix wan i2v pipeline bugs (#10975)

Álvaro Somoza (1):
      [Training] Better image interpolation in training scripts (#11206)

Author: fenrus75

Makefile

@@ -1,5 +1,5 @@
 PKG_NAME := pypi-diffusers
-URL = https://files.pythonhosted.org/packages/ee/72/d3f715e71a77f14a48f1ac081cea5bb6c6e6cd41c55b1291f401b1504679/diffusers-0.32.2.tar.gz
+URL = https://files.pythonhosted.org/packages/c2/4a/ea270e695ce5ffbffab2fdd542e8eba138d7cd26b945b6323a7791128e1b/diffusers-0.33.0.tar.gz
 ARCHIVES = 
 
 include ../common/Makefile.common

options.conf

@@ -1,6 +1,6 @@
 [package]
 name = pypi-diffusers
-url = https://files.pythonhosted.org/packages/ee/72/d3f715e71a77f14a48f1ac081cea5bb6c6e6cd41c55b1291f401b1504679/diffusers-0.32.2.tar.gz
+url = https://files.pythonhosted.org/packages/c2/4a/ea270e695ce5ffbffab2fdd542e8eba138d7cd26b945b6323a7791128e1b/diffusers-0.33.0.tar.gz
 archives = 
 giturl = https://github.com/huggingface/diffusers/
 domain = 

pypi-diffusers.spec

@@ -2,14 +2,14 @@
 # This file is auto-generated. DO NOT EDIT
 # Generated by: autospec.py
 # Using build pattern: pyproject
-# autospec version: v21
-# autospec commit: f4a13a5
+# autospec version: v24
+# autospec commit: a88ffdc
 #
 Name     : pypi-diffusers
-Version  : 0.32.2
-Release  : 64
-URL      : https://files.pythonhosted.org/packages/ee/72/d3f715e71a77f14a48f1ac081cea5bb6c6e6cd41c55b1291f401b1504679/diffusers-0.32.2.tar.gz
-Source0  : https://files.pythonhosted.org/packages/ee/72/d3f715e71a77f14a48f1ac081cea5bb6c6e6cd41c55b1291f401b1504679/diffusers-0.32.2.tar.gz
+Version  : 0.33.0
+Release  : 65
+URL      : https://files.pythonhosted.org/packages/c2/4a/ea270e695ce5ffbffab2fdd542e8eba138d7cd26b945b6323a7791128e1b/diffusers-0.33.0.tar.gz
+Source0  : https://files.pythonhosted.org/packages/c2/4a/ea270e695ce5ffbffab2fdd542e8eba138d7cd26b945b6323a7791128e1b/diffusers-0.33.0.tar.gz
 Summary  : State-of-the-art diffusion in PyTorch and JAX.
 Group    : Development/Tools
 License  : Apache-2.0
@@ -73,18 +73,18 @@ python3 components for the pypi-diffusers package.
 
 
 %prep
-%setup -q -n diffusers-0.32.2
-cd %{_builddir}/diffusers-0.32.2
+%setup -q -n diffusers-0.33.0
+cd %{_builddir}/diffusers-0.33.0
 pushd ..
-cp -a diffusers-0.32.2 buildavx2
+cp -a diffusers-0.33.0 buildavx2
 popd
 
 %build
 export http_proxy=http://127.0.0.1:9/
 export https_proxy=http://127.0.0.1:9/
 export no_proxy=localhost,127.0.0.1,0.0.0.0
 export LANG=C.UTF-8
-export SOURCE_DATE_EPOCH=1736963299
+export SOURCE_DATE_EPOCH=1744215018
 export GCC_IGNORE_WERROR=1
 export AR=gcc-ar
 export RANLIB=gcc-ranlib

release

@@ -1 +1 @@
-64
+65

upstream

@@ -1 +1 @@
-2172cdefb31f3b08fc81e9cb65597b14bed464ec/diffusers-0.32.2.tar.gz
+e40177300c8b3116b965abd524aed088939a53fc/diffusers-0.33.0.tar.gz