Add (Scheduled) Pseudo-Huber Loss training scripts to research projects (#7527)

* add scheduled pseudo-huber loss training scripts

See #7488

* add reduction modes to huber loss

* [DB Lora] *2 multiplier to huber loss cause of 1/2 a^2 conv.

pairing of kohya-ss/sd-scripts@c6495de

* [DB Lora] add option for smooth l1 (huber / delta)

Pairing of kohya-ss/sd-scripts@dd22958

* [DB Lora] unify huber scheduling

Pairing of kohya-ss/sd-scripts@19a834c

* [DB Lora] add snr huber scheduler

Pairing of kohya-ss/sd-scripts@47fb1a6

* fixup examples link

* use snr schedule by default in DB

* update all huber scripts with snr

* code quality

* huber: make style && make quality

---------

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

Author: kabachuha

examples/research_projects/scheduled_huber_loss_training/README.md

@@ -0,0 +1,15 @@
+# Scheduled Pseudo-Huber Loss for Diffusers
+
+These are the modifications of to include the possibility of training text2image models with Scheduled Pseudo Huber loss, introduced in https://arxiv.org/abs/2403.16728. (https://github.com/kabachuha/SPHL-for-stable-diffusion)
+
+## Why this might be useful?
+
+- If you suspect that the part of the training dataset might be corrupted, and you don't want these outliers to distort the model's supposed output
+
+- If you want to improve the aesthetic quality of pictures by helping the model disentangle concepts and be less influenced by another sorts of pictures.
+
+See https://github.com/huggingface/diffusers/issues/7488 for the detailed description.
+
+## Instructions
+
+The same usage as in the case of the corresponding vanilla Diffusers scripts https://github.com/huggingface/diffusers/tree/main/examples