GitLab

This is a better method than comparing against a list of supported backends as it allows for supporting any number of backends provided they are installed on the user's system.
This should have no effect on the behaviour of tests in Huggingface's CI workers.
See transformers#25506 where this approach has already been added.

default to None when fc alphas are not available.

* debugging starts

* debugging

* debugging

* debugging

* debugging

* debugging

* debugging ends, but does it?

* more robustness.

* enable lora for sdxl controlnets too.

* add: tests

* fix: assertion values.

* from audioldm

* unet down + mid

* vae, clap, flan-t5

* start sequence audio mae

* iterate on audioldm encoder

* finish encoder

* finish weight conversion

* text pre-processing

* gpt2 pre-processing

* fix projection model

* working

* unet equivalence

* finish in base

* add unet cond

* finish unet

* finish custom unet

* start clean-up

* revert base unet changes

* refactor pre-processing

* tests: from audioldm

* fix some tests

* more fixes

* iterate on tests

* make fix copies

* harden fast tests

* slow integration tests

* finish tests

* update checkpoint

* update copyright

* docs

* remove outdated method

* add docstring

* make style

* remove decode latents

* enable cpu offload

* (text_encoder_1, tokenizer_1) -> (text_encoder, tokenizer)

* more clean up

* more refactor

* build pr docs

* Update docs/source/en/api/pipelines/audioldm2.md
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

* small clean

* tidy conversion

* update for large checkpoint

* generate -> generate_language_model

* full clap model

* shrink clap-audio in tests

* fix large integration test

* fix fast tests

* use generation config

* make style

* update docs

* finish docs

* finish doc

* update tests

* fix last test

* syntax

* finalise tests

* refactor projection model in prep for TTS

* fix fast tests

* style

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

* Update loaders.py

add config_file to from_single_file, 
when the download_from_original_stable_diffusion_ckpt use

* Update loaders.py

add config_file to from_single_file,
when the download_from_original_stable_diffusion_ckpt use

* change config_file to original_config_file

* make style && make quality

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

* Impl tae slicing and tiling

* add tae tiling test

* add parameterized test

* formatted code

* fix failed test

* style docs

Fix `use_onnx` parameter usage in `from_pretrained` func and update `test_download_no_onnx_by_default` test (#4508)

* add missing use_onnx in from_pretrained func

* fix test_download_no_onnx_by_default test func

* address comments

* split test cases

* make safetensors default

* set default save method as safetensors

* update tests

* update to support saving safetensors

* update test to account for safetensors default

* update example tests to use safetensors

* update example to support safetensors

* update unet tests for safetensors

* fix failing loader tests

* fix qc issues

* fix pipeline tests

* fix example test

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

* Support higher dimension LoRAs

* add: tests

* fix: assertion values.

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

* fixed

* add: tests

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

fix loading custom text encoder when using from_single_file

* push to hub docs

* fix typo

* feedback

* make style

* core: add multicontrolnet support to sdxl controlnet

* modify checks.

* fix: original_size determination

* add: tests for multi controlnet sdxl.

* remove unnecessary prints.

* Fix unipc karras sigmas exception - fixes huggingface/diffusers#4580

* Add unipc scheduler tests for karras sigmas

* Add GLIGEN implementation

* GLIGEN: Fix code quality check failures

* GLIGEN: Fix Import block un-sorted or un-formatted failures

* GLIGEN: Fix check_repository_consistency failures

* GLIGEN: Add 'PositionNet' to versatile_diffusion/modeling_text_unet.py

* GLIGEN: check_repository_consistency: fix 'copy does not match' error

* GLIGEN: Fix review comments (1)

* GLIGEN: Fix E721 Do not compare types, use `isinstance()` failures

* GLIGEN : Ensure _encode_prompt() copy matches to StableDiffusionPipeline

* GLIGEN: Fix ruff E721 failure in unidiffuser/test_unidiffuser.py

* GLIGEN: doc_builder: restyle pipeline_stable_diffusion_gligen.py

* GIGLEN: reset files unrelated to gligen

* GLIGEN: Fix documentation comments (1)

* GLIGEN: Fix review comments (2)

* GLIGEN: Added FastTest

* GLIGEN: Fix review comments (3)

* add: pushtohubmixin to pipelines and schedulers docs overview.

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

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

[Pipeline utils] feat: implement push_to_hub for standalone models, schedulers as well as pipelines (#4128)

* feat: implement push_to_hub for standalone models.

* address PR feedback.

* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* remove max_shard_size.

* add: support for scheduler push_to_hub

* enable push_to_hub support for flax schedulers.

* enable push_to_hub for pipelines.

* Apply suggestions from code review
Co-authored-by: Lucain <lucainp@gmail.com>

* reflect pr feedback.

* address another round of deedback.

* better handling of kwargs.

* add: tests

* Apply suggestions from code review
Co-authored-by: Lucain <lucainp@gmail.com>

* setting hub staging to False for now.

* incorporate staging test as a separate job.
Co-authored-by: ydshieh <2521628+ydshieh@users.noreply.github.com>

* fix: tokenizer loading.

* fix: json dumping.

* move is_staging_test to a better location.

* better treatment to tokens.

* define repo_id to better handle concurrency

* style

* explicitly set token

* Empty-Commit

* move SUER, TOKEN to test

* collate org_repo_id

* delete repo

---------
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Lucain <lucainp@gmail.com>
Co-authored-by: ydshieh <2521628+ydshieh@users.noreply.github.com>

* add: utility for image grid.

* add: return type.

* change necessary places.

* add to utility page.

fix

* Fixing repo_id regex validation error on windows platforms

* Validating correct URL with prefix is provided

If we are loading a URL then we don't need to use os.path.join and array slicing to split out a repo_id and file path from an absolute filepath. 

Checking if the URL prefix is valid first before doing any URL splitting otherwise we raise a ValueError since neither a valid filepath or URL was provided.

* Style fixes

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

* [WIP] Remove code snippets containing `is_safetensors_available()`

* Modifying `import_utils.py`

* update pipeline tests for safetensor default

* fix test related to cached requests

* address import nits

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

* remove attention slicing

* apply feedback

* rename

* style

---------
Co-authored-by: yiyixuxu <yixu310@gmail,com>

* clean scheduler mixin

* up to dpmsolvermultistep

* finish cleaning

* first draft

* fix overview table

* apply feedback

* update reference code

* pin ruff version for quality checks

* update dependency versions table

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

* add load_lora_weights and save_lora_weights to StableDiffusionXLImg2ImgPipeline

* add load_lora_weights and save_lora_weights to StableDiffusionXLInpaintPipeline

* apply black format

* apply black format

* add copy statement

* fix statements

* fix statements

* fix statements

* run `make fix-copies`

Fix typos

* grad checkpointing

* fix make fix-copies

* fix

---------
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* correct

* correct blocks

* finish

* finish

* finish

* Apply suggestions from code review

* fix

* up

* up

* up

* Update examples/dreambooth/README_sdxl.md
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

* Apply suggestions from code review

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

make style
Co-authored-by: yiyixuxu <yixu310@gmail,com>

* add xformer

* enable_sequential_cpu_offload

* style

* Update src/diffusers/pipelines/kandinsky/pipeline_kandinsky_combined.py
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

---------
Co-authored-by: yiyixuxu <yixu310@gmail,com>
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

delete the duplicated code from the controlnet-img-2-img pipelines

Fix the typo that causes `omegaconf.errors.ConfigAttributeError: Missing key parms`

Allow controlnets to be loaded (from ckpt) in a parallel thread with a SD model (ckpt), and speed it up slightly (#4298)

Faster controlnet model instantiation, and allow controlnets to be loaded (from ckpt) in a parallel thread with a SD model (ckpt) without  tensor errors (race condition)

Fix typerror in pipeline handling for MultiControlNets which only contain a single ControlNet (#4454)

* Handle single controlnet in multicontrolnet

* Fix formatting