Add Community Example Consistency Training Script (#6717)

* initial commit for unconditional/class-conditional consistency training script

* make style

* Add entry for consistency training script in community README.

* Move consistency training script from community to research_projects/consistency_training

* Add requirements.txt and README to research_projects/consistency_training directory.

* Manually revert community README changes for consistency training.

* Fix path to script after moving script to research projects.

* Add option to load U-Net weights from pretrained model.

---------

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

examples/research_projects/consistency_training/README.md

@@ -0,0 +1,24 @@
+# Consistency Training
+
+`train_cm_ct_unconditional.py` trains a consistency model (CM) from scratch following the consistency training (CT) algorithm introduced in [Consistency Models](https://arxiv.org/abs/2303.01469) and refined in [Improved Techniques for Training Consistency Models](https://arxiv.org/abs/2310.14189). Both unconditional and class-conditional training are supported.
+
+A usage example is as follows:
+
+```bash
+accelerate launch examples/research_projects/consistency_training/train_cm_ct_unconditional.py \
+    --dataset_name="cifar10" \
+    --dataset_image_column_name="img" \
+    --output_dir="/path/to/output/dir" \
+    --mixed_precision=fp16 \
+    --resolution=32 \
+    --max_train_steps=1000 --max_train_samples=10000 \
+    --dataloader_num_workers=8 \
+    --noise_precond_type="cm" --input_precond_type="cm" \
+    --train_batch_size=4 \
+    --learning_rate=1e-04 --lr_scheduler="constant" --lr_warmup_steps=0 \
+    --use_8bit_adam \
+    --use_ema \
+    --validation_steps=100 --eval_batch_size=4 \
+    --checkpointing_steps=100 --checkpoints_total_limit=10 \
+    --class_conditional --num_classes=10 \
+```

examples/research_projects/consistency_training/requirements.txt

@@ -0,0 +1,6 @@
+accelerate>=0.16.0
+torchvision
+transformers>=4.25.1
+ftfy
+tensorboard
+Jinja2