obscura/diffusers
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
5,896 Commits 711 Branches 93 Tags
version-checks-cache
Commit Graph

5896 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Zijian Zhou d06750a5fd Fix autoencoder_kl_wan.py bugs for Wan2.2 VAE (#12335) 
* Update autoencoder_kl_wan.py

When using the Wan2.2 VAE, the spatial compression ratio calculated here is incorrect. It should be 16 instead of 8. Pass it in directly via the config to ensure it’s correct here.

* Update autoencoder_kl_wan.py
 2025-09-16 13:43:15 -10:00
Sari Hleihil 8c72cd12ee Added LucyEditPipeline (#12340) 
* Added LucyEditPipeline

* add import & stype

missing copied from

* Fix example doc string

---------

Co-authored-by: yiyixuxu <yixu310@gmail.com>
 2025-09-16 13:41:05 -10:00
Samarth Agrawal 751e250f70 fixed bug in defining embed dim for UNet1D (#12111) 
* fixed bug in defining embed dim

* matched 1d temb process to 2d

* Update src/diffusers/models/unets/unet_1d.py

Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>

---------

Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>
 2025-09-16 12:18:48 +05:30
Linoy Tsaban b50014067d Add Wan2.2 VACE - Fun (#12324) 
* support Wan2.2-VACE-Fun-A14B

* support Wan2.2-VACE-Fun-A14B

* support Wan2.2-VACE-Fun-A14B

* Apply style fixes

* test

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
 2025-09-15 21:31:26 +05:30
Daniel Socek f5c113e439 Use SDP on BF16 in GPU/HPU migration (#12310) 
* Use SDP on BF16 in GPU/HPU migration

Signed-off-by: Daniel Socek <daniel.socek@intel.com>

* Formatting fix for enabling SDP with BF16 precision on HPU

Signed-off-by: Daniel Socek <daniel.socek@intel.com>

---------

Signed-off-by: Daniel Socek <daniel.socek@intel.com>
 2025-09-12 08:00:36 -10:00
Sayak Paul 5e181eddfe Deprecate slicing and tiling methods from DiffusionPipeline (#12271) 
* deprecate slicing from flux pipeline.

* propagate.

* tiling

* up

* up
 2025-09-11 10:04:35 +05:30
Justin Ruan 55f0b3d758 Fix AttributeError of VisualClozeProcessor (#12121) 
Co-authored-by: YiYi Xu <yixu310@gmail.com>
 2025-09-11 04:17:34 +05:30
Sayak Paul eb7ef26736 [quant] allow components_to_quantize to be a non-list for single components (#12234) 
* allow non list components_to_quantize.

* up

* Apply suggestions from code review

* Apply suggestions from code review

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

* [docs] components_to_quantize (#12287)

init

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

---------

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
 2025-09-10 09:47:08 -10:00
ttio2tech e1b7f1f240 fix for the qwen controlnet pipeline - wrong device can be used (#12309) 
fix the device for textencoder
 2025-09-10 08:59:08 -10:00
Sayak Paul 9e7ae568d6 [feat] cache allocator warmup for from_single_model (#12305) 
* add

* add a test
 2025-09-10 12:55:32 +05:30
Sayak Paul f7b79452b4 [modular] fix flux modular pipelines for t2i and i2i (#12272) 
fix flux modular pipelines for t2i and i2i
 2025-09-10 12:39:55 +05:30
Sayak Paul 43459079ab [core] feat: support group offloading at the pipeline level (#12283) 
* feat: support group offloading at the pipeline level.

* add tests

* up

* [docs] Pipeline group offloading (#12286)

init

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

---------

Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
 2025-09-10 09:09:57 +05:30
kaixuanliu 4067d6c4b6 adjust criteria for marigold-intrinsics example on XPU (#12290) 
adjust criteria for XPU

Signed-off-by: Liu, Kaixuan <kaixuan.liu@intel.com>
Co-authored-by: Aryan <aryan@huggingface.co>
 2025-09-10 03:06:03 +05:30
calcuis 28106fcac4 gguf new quant type support (with demo) (#12076) 
* Update utils.py

not perfect but works

engine:
https://github.com/calcuis/gguf-connector/blob/main/src/gguf_connector/quant2c.py


inference example(s):
https://github.com/calcuis/gguf-connector/blob/main/src/gguf_connector/k6.py
https://github.com/calcuis/gguf-connector/blob/main/src/gguf_connector/k5.py

gguf file sample(s):
https://huggingface.co/calcuis/kontext-gguf/tree/main
https://huggingface.co/calcuis/krea-gguf/tree/main

* Apply style fixes

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
 2025-09-09 17:10:21 +05:30
Leo Jiang c222570a9b DeepSpeed adaption for flux-kontext (#12240) 
Co-authored-by: J石页 <jiangshuo9@h-partners.com>
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2025-09-09 12:58:08 +05:30
Frank (Haofan) Wang 4e36bb0d23 Support ControlNet-Inpainting for Qwen-Image (#12301) 
* add qwen-image-cn-inpaint


---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: yiyixuxu <yixu310@gmail.com>
 2025-09-08 14:59:26 -10:00
YiYi Xu f50b18eec7 [Modular] Qwen (#12220) 
* add qwen modular
 2025-09-08 00:27:02 -10:00
Steven Liu fc337d5853 [docs] Models (#12248) 
* init

* fix

* feedback

* feedback
 2025-09-05 11:52:09 -07:00
Steven Liu 32798bf242 [docs] Inference section cleanup (#12281) 
init

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2025-09-05 09:34:37 -07:00
Steven Liu c2e5ece08b [docs] Sharing pipelines/models (#12280) 
init
 2025-09-04 11:43:47 -07:00
co63oc 764b62473a fix some typos (#12265) 
Signed-off-by: co63oc <co63oc@users.noreply.github.com>
 2025-09-03 21:28:24 +05:30
Ju Hoon Park 6682956333 Add AttentionMixin to WanVACETransformer3DModel (#12268) 
* Add AttentionMixin to WanVACETransformer3DModel

to enable methods like `set_attn_processor()`.

* Import AttentionMixin in transformer_wan_vace.py

Special thanks to @tolgacangoz 🙇‍♂️
 2025-09-03 15:05:41 +05:30
Sayak Paul ffc8c0c1e1 [tests] feat: add AoT compilation tests (#12203) 
* feat: add a test for aot.

* up
 2025-09-03 11:15:27 +05:30
Ishan Modi 4acbfbf13b [Quantization] Add TRT-ModelOpt as a Backend (#11173) 
* initial commit

* update

* updates

* update

* update

* update

* update

* update

* update

* addressed PR comments

* update

* addressed PR comments

* update

* update

* update

* update

* update

* update

* updates

* update

* update

* addressed PR comments

* updates

* code formatting

* update

* addressed PR comments

* addressed PR comments

* addressed PR comments

* addressed PR comments

* fix docs and dependencies

* fixed dependency test

---------

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2025-09-03 10:14:52 +05:30
Steven Liu 6549b04ec6 [docs] AutoPipeline (#12160) 
* refresh

* feedback

* feedback

* supported models

* fix
 2025-09-02 21:06:26 -07:00
Sayak Paul 130fd8df54 [core] use kernels to support _flash_3_hub attention backend (#12236) 
* feat: try loading fa3 using kernels when available.

* up

* change to Hub.

* up

* up

* up

* switch env var.

* up

* up

* up

* up

* up

* up
 2025-09-03 08:48:07 +05:30
Dhruv Nair bcd4d77ba6 [CI] Remove big accelerator requirements from Quanto Tests (#12266) 
update
 2025-09-03 08:29:31 +05:30
Linoy Tsaban 006d092751 [Flux LoRA] fix for prior preservation and mixed precision sampling, follow up on #11873 (#12264) 
* propagate fixes from https://github.com/huggingface/diffusers/pull/11873/ to flux script

* propagate fixes from https://github.com/huggingface/diffusers/pull/11873/ to flux script

* propagate fixes from https://github.com/huggingface/diffusers/pull/11873/ to flux script

* Apply style fixes

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
 2025-09-02 11:30:33 +03:00
Ziheng Zhang 9e4a75b142 [docs] Fix VAE scale factor calculation in distributed inference docs (#12259) 
docs: Fix VAE scale factor calculation
 2025-09-01 16:34:16 -10:00
Bulat Akhmatov 0ff1aa910c [fix] fix for prior preservation and mixed precision sampling (#11873) 
Co-authored-by: Linoy Tsaban <57615435+linoytsaban@users.noreply.github.com>
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2025-09-01 14:12:14 +03:00
apolinário 901da9dccc Fix lora conversion function for ai-toolkit Qwen Image LoRAs (#12261) 
* Fix lora conversion function for ai-toolkit Qwen Image LoRAs

* add forgotten parenthesis

* remove space new line

* update pipeline

* detect if arrow or letter

* remove whitespaces

* style

* apply suggestion

* apply suggestion

* apply suggestion

---------

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2025-09-01 14:24:38 +05:30
Nguyễn Trọng Tuấn 67ffa7031e Add Qwen-Image-Edit Inpainting pipeline (#12225) 
* add qwenimage-edit inpaint feature

* stay up to date with main branch

* fix style

* fix docs

* copies

* fix

* again

* copies

---------

Co-authored-by: “Trgtuan10” <“tuannguyentrong.402@gmail.com”>
Co-authored-by: TuanNT-ZenAI <tuannt.zenai@gmail.com>
Co-authored-by: yiyixuxu <yixu310@gmail.com>
 2025-08-30 19:49:15 -10:00
Leo Jiang 827fad66a0 Improve performance of NPU FA (#12260) 
Co-authored-by: J石页 <jiangshuo9@h-partners.com>
Co-authored-by: Aryan <aryan@huggingface.co>
 2025-08-31 01:48:51 +05:30
Nguyễn Trọng Tuấn 9b721db205 [QwenImageEditPipeline] Add image entry in __call__ function (#12254) 
add  entry

Co-authored-by: TuanNT-ZenAI <tuannt.zenai@gmail.com>
 2025-08-29 20:16:43 -10:00
Dhruv Nair ba0e732eb0 [Modular] Consolidate load_default_components into load_components (#12217) 
* update

* Apply style fixes

* update

* update

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
 2025-08-28 19:55:02 +05:30
Dhruv Nair b2da59b197 [Modular] Provide option to disable custom code loading globally via env variable (#12177) 
* update

* update

* update

* update
 2025-08-28 19:54:32 +05:30
Dhruv Nair 7aa6af1138 [Refactor] Move testing utils out of src (#12238) 
* update

* update

* update

* update

* update

* merge main

* Revert "merge main"

This reverts commit 65efbcead5.
 2025-08-28 19:53:02 +05:30
Aryan 87b800e154 [modular diffusers] Fix AutoGuidance validation (#12247) 
fix
 2025-08-28 15:23:26 +05:30
YiYi Xu e58711e73c [Modular] support standard repo (#11944) 
* make modular pipeline work with model_index.json

* up

* style

* up

* up

* style

* up more

* Fix MultiControlNet import (#12118)

fix

---------

Co-authored-by: Álvaro Somoza <asomoza@users.noreply.github.com>
Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>
 2025-08-28 10:18:07 +02:00
Steven Liu cbecc33570 [docs] Reproducibility (#12237) 
* init

* dupe

* feedback
 2025-08-27 11:35:31 -07:00
Steven Liu 5237a82a35 [docs] Remove Flax (#12244) 
* remove flax

* toctree

* feedback
 2025-08-27 11:11:07 -07:00
Parag Ekbote 513dbdb2f3 Fix Table Rendering in ReadME (#12245) 
fix table rendering readme issue in readme.
 2025-08-27 10:44:49 -07:00
YiYi Xu 865ba102b3 [Qwen-Image] adding validation for guidance_scale, true_cfg_scale and negative_prompt (#12223) 
* up
 2025-08-27 01:04:33 -10:00
Manith Ratnayake 552c127c05 docs: correct typos in using-diffusers/other-formats (#12243) 2025-08-26 08:48:05 -07:00
Tianqi Tang 4b7fe044e3 Fix typos and inconsistencies (#12204) 
Fix typos and test assertions

Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
 2025-08-26 07:58:08 -07:00
Sayak Paul 532f41c999 Deprecate Flax support (#12151) 
* start removing flax stuff.

* add deprecation warning.

* add warning messages.

* more warnings.

* remove dockerfiles.

* remove more.

* Update src/diffusers/models/attention_flax.py

Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>

* up

---------

Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>
 2025-08-26 09:58:16 +02:00
Tolga Cangöz 5fcd5f560f Propose to update & upgrade SkyReels-V2 (#12167) 
* fix: update SkyReels-V2 documentation and moving into attn dispatcher

* Refactors SkyReelsV2's attention implementation

* style

* up

* Fixes formatting in SkyReels-V2 documentation

Wraps the visual demonstration section in a Markdown code block.

This change corrects the rendering of ASCII diagrams and examples, improving the overall readability of the document.

* Docs: Condense example arrays in skyreels_v2 guide

Improves the readability of the `step_matrix` examples by replacing long sequences of repeated numbers with a more compact `value×count` notation.

This change makes the underlying data patterns in the examples easier to understand at a glance.

* Add _repeated_blocks attribute to SkyReelsV2Transformer3DModel

* Refactor rotary embedding calculations in SkyReelsV2 to separate cosine and sine frequencies

* Enhance SkyReels-V2 documentation: update model loading for GPU support and remove outdated notes

* up

* up

* Update model_id in SkyReels-V2 documentation

* up

* refactor: remove device_map parameter for model loading and add pipeline.to("cuda") for GPU allocation

* fix: update copyright year to 2025 in skyreels_v2.md

* docs: enhance parameter examples and formatting in skyreels_v2.md

* docs: update example formatting and add notes on LoRA support in skyreels_v2.md

* refactor: remove copied comments from transformer_wan in SkyReelsV2 classes

* Clean up comments in skyreels_v2.md

Removed comments about acceleration helpers and Flash Attention installation.

* Add deprecation warning for `SkyReelsV2AttnProcessor2_0` class
 2025-08-26 12:54:19 +05:30
Leo Jiang 0fd7ee79ea NPU attention refactor for FLUX (#12209) 
* NPU attention refactor for FLUX transformer

* Apply style fixes

---------

Co-authored-by: J石页 <jiangshuo9@h-partners.com>
Co-authored-by: Aryan <aryan@huggingface.co>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
 2025-08-26 12:53:55 +05:30
sqt 0d1c5b0c3e Fix typo: 'will ge generated' -> 'will be generated' (#12231) 2025-08-25 12:47:52 -07:00
Meta 0e46c55931 Update README.md (#12193) 2025-08-25 11:35:56 -07:00