The dataset was generated using the code below:
import PIL
import torch
from datasets import Dataset, Features
from datasets import Image as ImageFeature
from datasets import Value, load_dataset
print("Loading pipeline...")
pipe_prior = DiffusionPipeline.from_pretrained(… See the full description on the dataset page: https://huggingface.co/datasets/diffusers-parti-prompts/kandinsky-2-2.
