GitLab

* WIP: flax FlaxDiffusionPipeline & FlaxStableDiffusionPipeline

* todo comment

* Fix imports

* Fix imports

* add dummies

* Fix empty init

* make pipeline work

* up

* Allow dtype to be overridden on model load.

This may be a temporary solution until #567 is addressed.

* Convert params to bfloat16 or fp16 after loading.

This deals with the weights, not the model.

* Use Flax schedulers (typing, docstring)

* PNDM: replace control flow with jax functions.

Otherwise jitting/parallelization don't work properly as they don't know
how to deal with traced objects.

I temporarily removed `step_prk`.

* Pass latents shape to scheduler set_timesteps()

PNDMScheduler uses it to reserve space, other schedulers will just
ignore it.

* Wrap model imports inside availability checks.

* Optionally return state in from_config.

Useful for Flax schedulers.

* Do not convert model weights to dtype.

* Re-enable PRK steps with functional implementation.

Values returned still not verified for correctness.

* Remove left over has_state var.

* make style

* Apply suggestion list -> tuple
Co-authored-by: Suraj Patil <surajp815@gmail.com>

* Apply suggestion list -> tuple
Co-authored-by: Suraj Patil <surajp815@gmail.com>

* Remove unused comments.

* Use zeros instead of empty.
Co-authored-by: Mishig Davaadorj <dmishig@gmail.com>
Co-authored-by: Mishig Davaadorj <mishig.davaadorj@coloradocollege.edu>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Suraj Patil <surajp815@gmail.com>

Signed-off-by: Ryan Russell <git@ryanrussell.org>
Signed-off-by: Ryan Russell <git@ryanrussell.org>

* Remove deprecated `torch_device` kwarg.

* Remove unused imports.

the link points to an old location of the train_unconditional.py file

* Return encoded texts by DiffusionPipelines

* Updated README to show hot to use enoded_text_input

* Reverted examples in README.md

* Reverted all

* Warning for long prompts

* Fix bugs

* Formatted

fix formula for noise levels in karras scheduler and tests

* docs: `src/diffusers` readability improvements
Signed-off-by: Ryan Russell <git@ryanrussell.org>

* docs: `make style` lint
Signed-off-by: Ryan Russell <git@ryanrussell.org>
Signed-off-by: Ryan Russell <git@ryanrussell.org>

Fix "ort is not defined" issue.

* refactor: pipelines readability improvements
Signed-off-by: Ryan Russell <git@ryanrussell.org>

* docs: remove todo comment from flax pipeline
Signed-off-by: Ryan Russell <git@ryanrussell.org>
Signed-off-by: Ryan Russell <git@ryanrussell.org>

the result of running the pipeline is stored in StableDiffusionPipelineOutput.images

* documenting `attention_flax.py` file

* documenting `embeddings_flax.py`

* documenting `unet_blocks_flax.py`

* Add new objs to doc page

* document `vae_flax.py`

* Apply suggestions from code review

* modify `unet_2d_condition_flax.py`

* make style

* Apply suggestions from code review

* make style

* Apply suggestions from code review

* fix indent

* fix typo

* fix indent unet

* Update src/diffusers/models/vae_flax.py

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

Signed-off-by: Ryan Russell <git@ryanrussell.org>

* Adding pred_original_sample to SchedulerOutput of DDPMScheduler, DDIMScheduler, LMSDiscreteScheduler, KarrasVeScheduler step methods so we can access the predicted denoised outputs

* Gave DDPMScheduler, DDIMScheduler and LMSDiscreteScheduler their own output dataclasses so the default SchedulerOutput in scheduling_utils does not need pred_original_sample as an optional extra

* Reordered library imports to follow standard

* didnt get import order quite right apparently

* Forgot to change name of LMSDiscreteSchedulerOutput

* Aha, needed some extra libs for make style to fully work

Signed-off-by: Ryan Russell <git@ryanrussell.org>
Signed-off-by: Ryan Russell <git@ryanrussell.org>

* add grad ckpt to downsample blocks

* make it work

* don't pass gradient_checkpointing to upsample block

* add tests for UNet2DConditionModel

* add test_gradient_checkpointing

* add gradient_checkpointing for up and down blocks

* add functions to enable and disable grad ckpt

* remove the forward argument

* better naming

* make supports_gradient_checkpointing private

* Handle the PIL.Image.Resampling deprecation

* style

Signed-off-by: Ryan Russell <git@ryanrussell.org>
Signed-off-by: Ryan Russell <git@ryanrussell.org>

* Optionally return state in from_config.

Useful for Flax schedulers.

* has_state is now a property, make check more strict.

I don't check the class is `SchedulerMixin` to prevent circular
dependencies. It should be enough that the class name starts with "Flax"
the object declares it "has_state" and the "create_state" exists too.

* Use state in pipeline from_pretrained.

* Make style

* Add smoke tests for the training examples

* upd

* use a dummy dataset

* mark as slow

* cleanup

* Update test cases

* naming

replace `dropout_prob` by `dropout` in `vae`

* Fix typo in docstring.

* Allow dtype to be overridden on model load.

This may be a temporary solution until #567 is addressed.

* Create latents in float32

The denoising loop always computes the next step in float32, so this
would fail when using `bfloat16`.

Fix params replication when sing the dummy checker.

* [Flax] Fix unet and ddim scheduler

* correct

* finish

* WIP: flax FlaxDiffusionPipeline & FlaxStableDiffusionPipeline

* todo comment

* Fix imports

* Fix imports

* add dummies

* Fix empty init

* make pipeline work

* up

* Use Flax schedulers (typing, docstring)

* Wrap model imports inside availability checks.

* more updates

* make sure flax is not broken

* make style

* more fixes

* up
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Pedro Cuenca <pedro@latenitesoft.com>

* use FlaxPreTrainedModel for flax safety module

* fix name

* fix one more

* Apply suggestions from code review

* Add the K-LMS scheduler to the inpainting pipeline + tests

* Remove redundant casts

* Fix BaseOutput initialization from dict

* style

* Simplify post-init, add tests

* remove debug

* rename weights to align with PT

* DiagonalGaussianDistribution => FlaxDiagonalGaussianDistribution

* fix name

* first commit:

- add `from_pt` argument in `from_pretrained` function
- add `modeling_flax_pytorch_utils.py` file

* small nit

- fix a small nit - to not enter in the second if condition

* major changes

- modify FlaxUnet modules
- first conversion script
- more keys to be matched

* keys match

- now all keys match
- change module names for correct matching
- upsample module name changed

* working v1

- test pass with atol and rtol= `4e-02`

* replace unsued arg

* make quality

* add small docstring

* add more comments

- add TODO for embedding layers

* small change

- use `jnp.expand_dims` for converting `timesteps` in case it is a 0-dimensional array

* add more conditions on conversion

- add better test to check for keys conversion

* make shapes consistent

- output `img_w x img_h x n_channels` from the VAE

* Revert "make shapes consistent"

This reverts commit 4cad1aeb4aeb224402dad13c018a5d42e96267f6.

* fix unet shape

- channels first!

* Fix a setting bug

* Fix typos

* Reverted params to parms

* Fix CrossAttention._sliced_attention
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>