GitLab

more elaborate example for peft torch.compile

* feat: add ip adapter benchmark

* sdxl support too.

* Empty-Commit

* initial

* check_inputs fix to the rest of pipelines

* add fix for no cfg too

* use of variable

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

* feat 256px diffusers inference bug

* change the max_length of T5 to pipeline config file

* fix bug in convert_pixart_alpha_to_diffusers.py

* Update scripts/convert_pixart_alpha_to_diffusers.py
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

* remove multi_scale_train parser

* Update src/diffusers/pipelines/pixart_alpha/pipeline_pixart_alpha.py
Co-authored-by: YiYi Xu <yixu310@gmail.com>

* Update src/diffusers/pipelines/pixart_alpha/pipeline_pixart_alpha.py
Co-authored-by: YiYi Xu <yixu310@gmail.com>

* styling

* change `model_token_max_length` to call argument.

* Refactoring

* add: max_sequence_length to the docstring.

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

* add: dreambooth lora script for Playground v2.5

* fix: kwarg

* address suraj's comments.

* Apply suggestions from code review
Co-authored-by: Suraj Patil <surajp815@gmail.com>

* apply suraj's suggestion

* incorporate changes in the canonical script./

* tracker naming

* fix: schedule determination

* add: two simple tests

* remove playground script

* note about edm-style training

* address pedro's comments.

* address part of Suraj's comments.

* Apply suggestions from code review
Co-authored-by: Suraj Patil <surajp815@gmail.com>

* remove guidance_scale.

* use mse_loss.

* add comments for preconditioning.

* quality

* Update examples/dreambooth/train_dreambooth_lora_sdxl.py
Co-authored-by: Suraj Patil <surajp815@gmail.com>

* tackle v-pred.

* Empty-Commit

* support edm for sdxl too.

* address suraj's comments.

* Empty-Commit

---------
Co-authored-by: Suraj Patil <surajp815@gmail.com>

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

* Speed up dataset mapping

* Fix missing columns

* Remove cache files cleanup

* Update examples/text_to_image/train_text_to_image_sdxl.py

* make style

* Fix code style

* style

* Empty-Commit

---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
Co-authored-by: Quentin Lhoest <42851186+lhoestq@users.noreply.github.com>
Co-authored-by: Quentin Lhoest <lhoest.q@gmail.com>

Update train_text_to_image_sdxl.py

update

* make workflow dispatchable.

* fix: lcm lora compile

pin uv version for now.

fix: path

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

---------
Co-authored-by: yiyixuxu <yixu310@gmail,com>
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

* Add copyright notice to relevant files and fix typos

* Set `timestep_spacing` parameter of `StableDiffusionXLPipeline`'s scheduler to `'trailing'`.

* Update `StableDiffusionXLPipeline.from_single_file` by including EulerAncestralDiscreteScheduler with `timestep_spacing="trailing"` param.

* Update model loading method in SDXL Turbo documentation

* update

* update

* use uv for installing stuff in the workflows.

* fix: from source installation command when using uv.

* fix uv venv issue

* edit editable installation.

* fix quality installation

* checking

* make editable.

* more check

* check

* add: export step

* venv handling.

* checking.

* fix: dependency workflows.

* peft tests.

* proper way to initialize env.

* Empty-Commit

* Empty-Commit

update

* move model helper function in pipeline to EfficiencyMixin

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

* fix

* fix

* remove stalebot's ability to close issues

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

update

* fix kwarg

* Empty-Commit

make tests conditional and other things.

* DPMMultistep rescale_betas_zero_snr

* DPM upcast samples in step()

* DPM rescale_betas_zero_snr UT

* DPMSolverMulti move sample upcast after model convert

Avoids having to re-use the dtype.

* Add a newline for Ruff

Fix typos, formatting and remove trailing whitespace

* add DPM scheduler with EDM formulation

* set sigmas in init

* add _compute_sigmas

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

* address some review comments

* up,

* add tests

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

* Add EDMEulerScheduler

* address review comments

* fix import

* fix test

* add tests

* add co-author

Co-authored-by:  @dg845 dgu8957@gmail.com

remove uv from docker jax tpu

* denormalize latents with the mean and std if available

* fix denormalize

* add latent mean and std in vae config

* address sayak's comment

* update

* update

* update

* update

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

move to uv in the Dockerfiles.

* log_validation unification for controlnet.

* additional fixes.

* remove print.

* better reuse and loading

* make final inference run conditional.

* Update examples/controlnet/README_sdxl.md
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>

* resize the control image in the snippet.

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

* pass revision in the loading_kwarhs.

* remove revision from load_sub_model.

* copy the state dict in load lora weights

* fixup

edit templates to remove patrick's name.

* Make LoRACompatibleConv padding_mode work.

* Format code style.

* add fast test

* Update src/diffusers/models/lora.py

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

* code refactor

* apply patrickvonplaten suggestion to simplify the code.

* rm test_lora_layers_old_backend.py and add test case in test_lora_layers_peft.py

* update test case.

---------
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: YiYi Xu <yixu310@gmail.com>

* Fix typos

* update

* make style