GitLab

create a utility for casting the lora params during training.

* fix: training resume from fp16.

* add: comment

* remove residue from another branch.

* remove more residues.

* thanks to Younes; no hacks.

* style.

* clean things a bit and modularize _set_state_dict_into_text_encoder

* add comment about the fix detailed.

* support compile

* make style

* move unwrap_model inside function

* change unwrap call

* run make style

* Update examples/dreambooth/train_dreambooth.py
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

* Revert "Update examples/dreambooth/train_dreambooth.py"

This reverts commit 70ab09732e7cfec0b19c497f823ddd1c8259dad0.

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

* added unwrapper

* fiz typo

* update: make controlnet script torch compile compatible
Signed-off-by: Suvaditya Mukherjee <suvadityamuk@gmail.com>

* update: correct earlier mistakes for compilation
Signed-off-by: Suvaditya Mukherjee <suvadityamuk@gmail.com>

* update: fix code style issues
Signed-off-by: Suvaditya Mukherjee <suvadityamuk@gmail.com>

---------
Signed-off-by: Suvaditya Mukherjee <suvadityamuk@gmail.com>

* make torch.compile compatible

* fix quality

support compile

pass tracker name as argumentw

* add stylealigned sdxl pipeline

* bugfix

* update docs

* remove einops dependency

* update README

* update example docstring

make checkpointing compatible when using torch.compile.

* Fix bug where unet's time_cond_proj_dim is not set correctly if using args.unet_time_cond_proj_dim.

* make style

* base template file - train_instruct_pix2pix.py

* additional import and parser argument requried for lora

* finetune only instructpix2pix model -- no need to include these layers

* inject lora layers

* freeze unet model -- only lora layers are trained

* training modifications to train only lora parameters

* store only lora parameters

* move train script to research project

* run quality and style code checks

* move train script to a new folder

* add README

* update README

* update references in README

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

* make it torch.compile comaptible

* make the text encoder compatible too.

* style

* upload codes and doc

* lint

* lint

* lint

* update code

* remove blank lines

* Fix load url

* enable stable-xl textual inversion

* check if optimizer_2 exists

* check text_encoder_2 before using

* add textual inversion for sdxl in a single file

* fix style

* fix example style

* reset for error changes

* add readme for sdxl

* fix style

* disable autocast as it will cause cast error when weight_dtype=bf16

* fix spelling error

* fix style and readme and 8bit optimizer

* add README_sdxl.md link

* add tracker key on log_validation

* run style

* rm the second center crop

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

* init works

* add gluegen pipeline

* add gluegen code

* add another way to load language adapter

* make style

* Update README.md

* change doc

* add: experimental script for diffusion dpo training.

* random_crop cli.

* fix: caption tokenization.

* fix: pixel_values index.

* fix: grad?

* debug

* fix: reduction.

* fixes in the loss calculation.

* style

* fix: unwrap call.

* fix: validation inference.

* add: initial sdxl script

* debug

* make sure images in the tuple are of same res

* fix model_max_length

* report print

* boom

* fix: numerical issues.

* fix: resolution

* comment about resize.

* change the order of the training transformation.

* save call.

* debug

* remove print

* manually detaching necessary?

* use the same vae for validation.

* add: readme.

* post release

* style

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

* null-text-inversion-implementation

* edited

* edited

* edited

* edited

* edited

* edit

* makestyle

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

* unwrap text encoder when saving hook only for full text encoder tuning

* unwrap text encoder when saving hook only for full text encoder tuning

* save embeddings in each checkpoint as well

* save embeddings in each checkpoint as well

* save embeddings in each checkpoint as well

* Update examples/advanced_diffusion_training/train_dreambooth_lora_sdxl_advanced.py
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

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

* Intel Gen 4 Xeon and later support bf16

* fix bf16 notes

* Make WDS pipeline interpolation type configurable.

* Make the VAE encoding batch size configurable.

* Make lora_alpha and lora_dropout configurable for LCM LoRA scripts.

* Generalize scalings_for_boundary_conditions function and make the timestep scaling configurable.

* Make LoRA target modules configurable for LCM-LoRA scripts.

* Move resolve_interpolation_mode to src/diffusers/training_utils.py and make interpolation type configurable in non-WDS script.

* apply suggestions from review

Update README_sdxl.md

* handle rest of the stuff related to deprecated lora stuff.

* fix: copies

* don't modify the uNet in-place.

* fix: temporal autoencoder.

* manually remove lora layers.

* don't copy unet.

* alright

* remove lora attn processors from unet3d

* fix: unet3d.

* styl

* Empty-Commit

* add clip_skip, freeu, qkv

* fix

* add ip-adapter support

* callback on step end

* update

* fix NoneType bug

* fix

* add guidance scale embedding

* add textual inversion

[bug fix] using snr gamma and prior preservation loss in the dreambooth lora sdxl training scripts (#6356)

* change timesteps used to calculate snr when --with_prior_preservation is enabled

* change timesteps used to calculate snr when --with_prior_preservation is enabled (canonical script)

* style

* revert canonical script to before snr gamma change

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

* Fix gradient-checkpointing option is ignored in SDXL+LoRA training. (#6388)

* Fix gradient-checkpointing option is ignored in SD+LoRA training.

* Fix gradient checkpoint is not applied to text encoders. (SDXL+LoRA)

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

* Add WebUI format support to Advanced Training Script

* style

---------
Co-authored-by: multimodalart <joaopaulo.passos+multimodal@gmail.com>

* files added

* fixing code quality

* fixing code quality

* fixing code quality

* fixing code quality

* sorted import block

* seperated import wandb

* ruff on script

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

* add to dreambooth lora.

* add: t2i lora.

* add: sdxl t2i lora.

* style

* lcm lora sdxl.

* unwrap

* fix: enable_adapters().

fix keys for lora format on advanced training scripts

* Fix ProdigyOPT in SDXL Dreambooth script

* style

* style

* Add PEFT to Advanced Training Script

* style

* style

* ✨ style ✨



* change order for logic operation

* add lora alpha

* style

* Align PEFT to new format

* Update train_dreambooth_lora_sdxl_advanced.py

Apply #6355 fix

---------
Co-authored-by: multimodalart <joaopaulo.passos+multimodal@gmail.com>

* fix

* style fix

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

Fix bug when creating the guidance embeddings using multiple GPUs.
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

* update

* update

* update

* update

* update

* make style

* remove docs

* update

* move to research folder.

* fix-copies

* remove _toctree entry.

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

* spit diffusers-native format from the get go.

* rejig the peft_to_diffusers mapping.

* amused update links to new repo

* lint