* update svd docs
* fix example doc string
* update return type hints/docs
* update type hints
* Fix typos in pipeline_stable_video_diffusion.py
* make style && make fix-copies
* Update src/diffusers/pipelines/stable_video_diffusion/pipeline_stable_video_diffusion.py
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
* Update src/diffusers/pipelines/stable_video_diffusion/pipeline_stable_video_diffusion.py
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
* update based on suggestion
---------
Co-authored-by: M. Tolga Cangöz <mtcangoz@gmail.com>
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
* Enable FakeTensorMode for EulerDiscreteScheduler scheduler
PyTorch's FakeTensorMode does not support `.numpy()` or `numpy.array()`
calls.
This PR replaces `sigmas` numpy tensor by a PyTorch tensor equivalent
Repro
```python
with torch._subclasses.FakeTensorMode() as fake_mode, ONNXTorchPatcher():
fake_model = DiffusionPipeline.from_pretrained(model_name, low_cpu_mem_usage=False)
```
that otherwise would fail with
`RuntimeError: .numpy() is not supported for tensor subclasses.`
* Address comments
* add tags for diffusers training
* add tags for diffusers training
* add tags for diffusers training
* add tags for diffusers training
* add tags for diffusers training
* add tags for diffusers training
* add dora tags for drambooth lora scripts
* style
* add is_dora arg
* style
* add dora training feature to sd 1.5 script
* added notes about DoRA training
---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
* 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>
* 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
* move model helper function in pipeline to EfficiencyMixin
---------
Co-authored-by: YiYi Xu <yixu310@gmail.com>
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
* 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
* 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>
* 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>
