Remove <cat-toy> validation prompt from textual_inversion_sdxl.py
The `<cat-toy>` validation prompt is a default choice for the example task in the README. But no other part of `textual_inversion_sdxl.py` references the cat toy and `textual_inversion.py` has a default validation prompt of `None` as well.
So bring `textual_inversion_sdxl.py` in line with `textual_inversion.py` and change default validation prompt to `None`
* attention_head_dim
* debug
* print more info
* correct num_attention_heads behaviour
* down_block_num_attention_heads -> num_attention_heads.
* correct the image link in doc.
* add: deprecation for num_attention_head
* fix: test argument to use attention_head_dim
* more fixes.
* quality
* address comments.
* remove depcrecation.
* add: support for passing ip adapter image embeddings
* debugging
* make feature_extractor unloading conditioned on safety_checker
* better condition
* type annotation
* index to look into value slices
* more debugging
* debugging
* serialize embeddings dict
* better conditioning
* remove unnecessary prints.
* Update src/diffusers/loaders/ip_adapter.py
Co-authored-by: YiYi Xu <yixu310@gmail.com>
* make fix-copies and styling.
* styling and further copy fixing.
* fix: check_inputs call in controlnet sdxl img2img pipeline
---------
Co-authored-by: YiYi Xu <yixu310@gmail.com>
* feat: standarize model card creation for dreambooth training.
* correct 'inference
* remove comments.
* take component out of kwargs
* style
* add: card template to have a leaner description.
* widget support.
* propagate changes to train_dreambooth_lora
* propagate changes to custom diffusion
* make widget properly type-annotated
* fix: callback function name is incorrect
On this tutorial there is a function defined and then used inside `callback_on_step_end` argument, but the name was not correct (mismatch)
* fix: typo in num_timestep (correct is num_timesteps)
fixed property name
* remove _to_tensor
* remove _to_tensor definition
* remove _collapse_frames_into_batch
* remove lora for not bloating the code.
* remove sample_size.
* simplify code a bit more
* ensure timesteps are always in tensor.
* Fix `AutoencoderTiny` with `use_slicing`
When using slicing with AutoencoderTiny, the encoder mistakenly encodes the entire batch for every image in the batch.
* Fixed formatting issue
* add noise_offset param
* micro conditioning - wip
* image processing adjusted and moved to support micro conditioning
* change time ids to be computed inside train loop
* change time ids to be computed inside train loop
* change time ids to be computed inside train loop
* time ids shape fix
* move token replacement of validation prompt to the same section of instance prompt and class prompt
* add offset noise to sd15 advanced script
* fix token loading during validation
* fix token loading during validation in sdxl script
* a little clean
* style
* a little clean
* style
* sdxl script - a little clean + minor path fix
sd 1.5 script - change default resolution value
* ad 1.5 script - minor path fix
* fix missing comma in code example in model card
* clean up commented lines
* style
* remove time ids computed outside training loop - no longer used now that we utilize micro-conditioning, as all time ids are now computed inside the training loop
* style
* [WIP] - added draft readme, building off of examples/dreambooth/README.md
* readme
* readme
* readme
* readme
* readme
* readme
* readme
* readme
* removed --crops_coords_top_left from CLI args
* style
* fix missing shape bug due to missing RGB if statement
* add blog mention at the start of the reamde as well
* Update examples/advanced_diffusion_training/README.md
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
* change note to render nicely as well
---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
Fix bug in ResnetBlock2D.forward when not USE_PEFT_BACKEND and using scale_shift for time emb where the lora scale gets overwritten.
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>
