GitLab

* [tests] use parent class for monkey patching to not break other tests

* fix

* add: test for testing unloading lora.

* add :reason to skipif.

* initial implementation of lora unload().

* apply styling.

* add: doc.

* change checkpoints.

* reinit generator

* finalize slow test.

* add fast test for unloading lora.

force accelerate to be installed

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

* community pipeline: implementation of iadb

* iadb.py: reformat using black

* iadb.py: linting update

* Multiply lr scheduler steps by `num_processes`.

* Stop multiplying steps by gradient accumulation.

* Finish

* make style

increase the time of timeout when using big dataset or high resolution

* Add circular padding option

* Fix style with black

* Fix corner case with small image size

* Add circular padding test cases

* Fix docstring

* Improve docstring for circular padding, remove slow test case

* Update docs for circular padding argument

* Add images comparison for circular padding

* diffusers#4003 - initial implementation of max_inference_steps

* diffusers#4003 - initial implementation of max_inference_steps and first_inference_step for img2img

* diffusers#4003 - use first_inference_step as an input arg for get_timestamps in img2img

* diffusers#4003 Do not add noise during img2img when we have a defined first timestep

* diffusers#4003 Mild updates after revert

* diffusers#4003 Missing change

* Show implementation with denoising_start and end

* Apply suggestions from code review

* Update src/diffusers/pipelines/stable_diffusion_xl/pipeline_stable_diffusion_xl.py
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>

* move to 0.19.0dev

* Apply suggestions from code review

* add exhaustive tests

* add docs

* finish

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

* make style

---------
Co-authored-by: bghira <bghira@users.github.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

* 📝 Fix broken link to models documentation

Corrected the link to the models documentation in the README. Previously, the link was pointing to an incorrect URL. Now, the link directs users to the correct documentation page for more details on the models.

Thanks! 🙌



* Update src/diffusers/models/README.md
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>

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

Fixed parameter descriptions

fix diffedit code mistake

typo fix in example docstring in audioldm pipeline

Signed-off-by: Raphael <oOraph@users.noreply.github.com>
Co-authored-by: Raphael <oOraph@users.noreply.github.com>

* start improving single file load

* Fix more

* start improving single file load

* Fix sd 2.1

* further improve from_single_file

FIX force_download in download utils

* feat: add a cli util for fp16 and safetensors format.

* fix: commit_description.

* add: usage example.

* add dreambooth lora script for SDXL incorporating latest changes.

* remove use_auth_token=True.

* add: documentation

* remove unneeded cli.

* increase the number of training steps in the readme.

* add LoraLoaderMixin to the subclassing mix.

* add sdxl lora dreambooth test.

* add: inference code sample.

* add: refiner output.

* add LoraLoaderMixin to the mix of classes of StableDiffusionXLImg2ImgPipeline.

* change default resolution of DreamBoothDataset.

* better sdxl report path.

* Apply suggestions from code review
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>

---------
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>

Update CONTRIBUTING.md

* fix

* correct link

* Fix typo

---------
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>

* improve loading subcomponents

* Add test for logging

* improve loading subcomponents

* make style

* make style

* fix

* finish

Remove remaining `not` in upscale pipeline.

* minor improvements to the SDXL doc.

* use_refiner variable.

* fix: typo.

* Update pipeline_stable_diffusion_xl.py

fix a bug

* Update pipeline_stable_diffusion_xl_img2img.py

* Update pipeline_stable_diffusion_xl_img2img.py

* Update pipeline_stable_diffusion_upscale.py

* style

---------
Co-authored-by: Hu Ye <xiaohuzc@gmail.com>

* Deprecate variant nicely

* make style

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

* Apply suggestions from code review
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>

---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>

* Fix all sequential offload

* make style

* make style

Update pipeline_dance_diffusion.py

* refactor to support patching LoRA into T5

instantiate the lora linear layer on the same device as the regular linear layer

get lora rank from state dict

tests

fmt

can create lora layer in float32 even when rest of model is float16

fix loading model hook

remove load_lora_weights_ and T5 dispatching

remove Unet#attn_processors_state_dict

docstrings

* text encoder monkeypatch class method

* fix test

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

* Add sdxl prompt embeddings

* Fix more

* fix some slow tests