GitLab

* use real instead of complex tensors in Wan2.1 RoPE

* remove the redundant type conversion

* unpack rotary_emb

* register rotary embedding frequencies as non-persistent buffers

* Apply style fixes

---------
Co-authored-by: Aryan <aryan@huggingface.co>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

* update

* update

* update docs

remove print

* ENH Improve speed of expanding LoRA scales

Resolves #11816

The following call proved to be a bottleneck when setting a lot of LoRA
adapters in diffusers:

https://github.com/huggingface/diffusers/blob/cdaf84a708eadf17d731657f4be3fa39d09a12c0/src/diffusers/loaders/peft.py#L482

This is because we would repeatedly call unet.state_dict(), even though
in the standard case, it is not necessary:

https://github.com/huggingface/diffusers/blob/cdaf84a708eadf17d731657f4be3fa39d09a12c0/src/diffusers/loaders/unet_loader_utils.py#L55



This PR fixes this by deferring this call, so that it is only run when
it's necessary, not earlier.

* Small fix

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

* feat: use exclude modules to loraconfig.

* version-guard.

* tests and version guard.

* remove print.

* describe the test

* more detailed warning message + shift to debug

* update

* update

* update

* remove test

* fix: lora unloading behvaiour

* fix

* update

* update

* add test

* address review comments

* update

* fixes

* change decorator order to fix tests

* try fix

* fight tests

fix

* support flux kontext

* make fix-copies

* add example

* add tests

* update docs

* update

* add note on integrity checker

* initial commit

* initial commit

* add readme section and fixes in the training script.

* add test

* rectify ckpt_id

* fix ckpt

* fixes

* change id

* update

* Update examples/dreambooth/train_dreambooth_lora_flux_kontext.py
Co-authored-by: Aryan <aryan@huggingface.co>

* Update examples/dreambooth/README_flux.md

---------
Co-authored-by: Aryan <aryan@huggingface.co>
Co-authored-by: linoytsaban <linoy@huggingface.co>
Co-authored-by: Linoy Tsaban <57615435+linoytsaban@users.noreply.github.com>

* support flux kontext

* make fix-copies

* add example

* add tests

* update docs

* update

* add note on integrity checker

* make fix-copies issue

* add copied froms

* make style

* update repository ids

* more copied froms

* [rfc][compile] compile method for DiffusionPipeline

* Apply suggestions from code review
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

* Apply style fixes

* Update docs/source/en/optimization/fp16.md

* check

---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

* update

* update

* update

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

* post release v0.34.0

* code quality

---------
Co-authored-by: YiYi Xu <yixu310@gmail.com>

* update

* update

* update

* raise as early as possible in group offloading

* remove check from ModuleGroup

up

* minor cleanups in the lora docs.

* Apply suggestions from code review
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

* format docs

* fix copies

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

enable cpu offloading of new pipelines on XPU & use device agnostic empty to make pipelines work on XPU (#11671)

* commit 1
Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* patch 2
Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* Update pipeline_pag_sana.py

* Update pipeline_sana.py

* Update pipeline_sana_controlnet.py

* Update pipeline_sana_sprint_img2img.py

* Update pipeline_sana_sprint.py

* fix style
Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* fix fat-thumb while merge conflict
Signed-off-by: YAO Matrix <matrix.yao@intel.com>

* fix ci issues
Signed-off-by: YAO Matrix <matrix.yao@intel.com>

---------
Signed-off-by: YAO Matrix <matrix.yao@intel.com>
Co-authored-by: Ilyas Moutawwakil <57442720+IlyasMoutawwakil@users.noreply.github.com>

Fix dimensionality in `apply_rotary_emb` functions' comments.

update

* start

* updates

* update

* update

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

* start implementing disk offloading in group.

* delete diff file.

* updates.patch

* offload_to_disk_path

* check if safetensors already exist.

* add test and clarify.

* updates

* update todos.

* update more docs.

* update docs

* factor out stuff from load_lora_adapter().

* simplifying text encoder lora loading.

* fix peft.py

* fix logging locations.

* formatting

* fix

* update

* update

* update

update

update

add is_compileable property to quantizers.

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* update

* updte

* update

* update

* update

change to 2025 licensing for remaining

* ⚡

️ Speed up method `AutoencoderKLWan.clear_cache` by 886%

**Key optimizations:**
- Compute the number of `WanCausalConv3d` modules in each model (`encoder`/`decoder`) **only once during initialization**, store in `self._cached_conv_counts`. This removes unnecessary repeated tree traversals at every `clear_cache` call, which was the main bottleneck (from profiling).
- The internal helper `_count_conv3d_fast` is optimized via a generator expression with `sum` for efficiency.

All comments from the original code are preserved, except for updated or removed local docstrings/comments relevant to changed lines.  
**Function signatures and outputs remain unchanged.**

* Apply style fixes

* Apply suggestions from code review
Co-authored-by: Aryan <contact.aryanvs@gmail.com>

* Apply style fixes

---------
Co-authored-by: codeflash-ai[bot] <148906541+codeflash-ai[bot]@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Aryan <aryan@huggingface.co>
Co-authored-by: Aryan <contact.aryanvs@gmail.com>
Co-authored-by: Aseem Saxena <aseem.bits@gmail.com>

update

* show how metadata stuff should be incorporated in training scripts.

* typing

* fix

---------
Co-authored-by: Linoy Tsaban <57615435+linoytsaban@users.noreply.github.com>

* fix flux lora loader when return_metadata is true for non-diffusers

* remove annotation

* working state from hameerabbasi and iddl

* working state form hameerabbasi and iddl (transformer)

* working state (normalization)

* working state (embeddings)

* add chroma loader

* add chroma to mappings

* add chroma to transformer init

* take out variant stuff

* get decently far in changing variant stuff

* add chroma init

* make chroma output class

* add chroma transformer to dummy tp

* add chroma to init

* add chroma to init

* fix single file

* update

* update

* add chroma to auto pipeline

* add chroma to pipeline init

* change to chroma transformer

* take out variant from blocks

* swap embedder location

* remove prompt_2

* work on swapping text encoders

* remove mask function

* dont modify mask (for now)

* wrap attn mask

* no attn mask (can't get it to work)

* remove pooled prompt embeds

* change to my own unpooled embeddeer

* fix load

* take pooled projections out of transformer

* ensure correct dtype for chroma embeddings

* update

* use dn6 attn mask + fix true_cfg_scale

* use chroma pipeline output

* use DN6 embeddings

* remove guidance

* remove guidance embed (pipeline)

* remove guidance from embeddings

* don't return length

* dont change dtype

* remove unused stuff, fix up docs

* add chroma autodoc

* add .md (oops)

* initial chroma docs

* undo don't change dtype

* undo arxiv change

unsure why that happened

* fix hf papers regression in more places

* Update docs/source/en/api/pipelines/chroma.md
Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>

* do_cfg -> self.do_classifier_free_guidance

* Update docs/source/en/api/models/chroma_transformer.md
Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>

* Update chroma.md

* Move chroma layers into transformer

* Remove pruned AdaLayerNorms

* Add chroma fast tests

* (untested) batch cond and uncond

* Add # Copied from for shift

* Update # Copied from statements

* update norm imports

* Revert cond + uncond batching

* Add transformer tests

* move chroma test (oops)

* chroma init

* fix chroma pipeline fast tests

* Update src/diffusers/models/transformers/transformer_chroma.py
Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>

* Move Approximator and Embeddings

* Fix auto pipeline + make style, quality

* make style

* Apply style fixes

* switch to new input ids

* fix # Copied from error

* remove # Copied from on protected members

* try to fix import

* fix import

* make fix-copes

* revert style fix

* update chroma transformer params

* update chroma transformer approximator init params

* update to pad tokens

* fix batch inference

* Make more pipeline tests work

* Make most transformer tests work

* fix docs

* make style, make quality

* skip batch tests

* fix test skipping

* fix test skipping again

* fix for tests

* Fix all pipeline test

* update

* push local changes, fix docs

* add encoder test, remove pooled dim

* default proj dim

* fix tests

* fix equal size list input

* update

* push local changes, fix docs

* add encoder test, remove pooled dim

* default proj dim

* fix tests

* fix equal size list input

* Revert "fix equal size list input"

This reverts commit 3fe4ad67d58d83715bc238f8654f5e90bfc5653c.

* update

* update

* update

* update

* update

---------
Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

* support text-to-image

* update example

* make fix-copies

* support use_flow_sigmas in EDM scheduler instead of maintain cosmos-specific scheduler

* support video-to-world

* update

* rename text2image pipeline

* make fix-copies

* add t2i test

* add test for v2w pipeline

* support edm dpmsolver multistep

* update

* update

* update

* update tests

* fix tests

* safety checker

* make conversion script work without guardrail

* feat: parse metadata from lora state dicts.

* tests

* fix tests

* key renaming

* fix

* smol update

* smol updates

* load metadata.

* automatically save metadata in save_lora_adapter.

* propagate changes.

* changes

* add test to models too.

* tigher tests.

* updates

* fixes

* rename tests.

* sorted.

* Update src/diffusers/loaders/lora_base.py
Co-authored-by: Benjamin Bossan <BenjaminBossan@users.noreply.github.com>

* review suggestions.

* removeprefix.

* propagate changes.

* fix-copies

* sd

* docs.

* fixes

* get review ready.

* one more test to catch error.

* change to a different approach.

* fix-copies.

* todo

* sd3

* update

* revert changes in get_peft_kwargs.

* update

* fixes

* fixes

* simplify _load_sft_state_dict_metadata

* update

* style fix

* uipdate

* update

* update

* empty commit

* _pack_dict_with_prefix

* update

* TODO 1.

* todo: 2.

* todo: 3.

* update

* update

* Apply suggestions from code review
Co-authored-by: Benjamin Bossan <BenjaminBossan@users.noreply.github.com>

* reraise.

* move argument.

---------
Co-authored-by: Benjamin Bossan <BenjaminBossan@users.noreply.github.com>
Co-authored-by: Linoy Tsaban <57615435+linoytsaban@users.noreply.github.com>