Datasets:

Dragonjinny
/

FiFA-pickapic-v2

pretty_name: Pick-a-Pic v2 · FiFA Filtered
license: mit
dataset_info:
  - config_name: FiFA-100k
    features:
      - name: are_different
        dtype: bool
      - name: best_image_uid
        dtype: string
      - name: caption
        dtype: string
      - name: created_at
        dtype: timestamp[ns]
      - name: has_label
        dtype: bool
      - name: image_0_uid
        dtype: string
      - name: image_0_url
        dtype: string
      - name: image_1_uid
        dtype: string
      - name: image_1_url
        dtype: string
      - name: jpg_0
        dtype: binary
      - name: jpg_1
        dtype: binary
      - name: label_0
        dtype: float64
      - name: label_1
        dtype: float64
      - name: model_0
        dtype: string
      - name: model_1
        dtype: string
      - name: ranking_id
        dtype: int64
      - name: user_id
        dtype: int64
      - name: num_example_per_prompt
        dtype: int64
      - name: __index_level_0__
        dtype: int64
    splits:
      - name: train
        num_bytes: 32217070275.579575
        num_examples: 95948
    download_size: 32218306878
    dataset_size: 32217070275.579575
  - config_name: FiFA-10k
    features:
      - name: are_different
        dtype: bool
      - name: best_image_uid
        dtype: string
      - name: caption
        dtype: string
      - name: created_at
        dtype: timestamp[ns]
      - name: has_label
        dtype: bool
      - name: image_0_uid
        dtype: string
      - name: image_0_url
        dtype: string
      - name: image_1_uid
        dtype: string
      - name: image_1_url
        dtype: string
      - name: jpg_0
        dtype: binary
      - name: jpg_1
        dtype: binary
      - name: label_0
        dtype: float64
      - name: label_1
        dtype: float64
      - name: model_0
        dtype: string
      - name: model_1
        dtype: string
      - name: ranking_id
        dtype: int64
      - name: user_id
        dtype: int64
      - name: num_example_per_prompt
        dtype: int64
      - name: __index_level_0__
        dtype: int64
    splits:
      - name: train
        num_bytes: 3218416419.2211432
        num_examples: 9585
    download_size: 3216732485
    dataset_size: 3218416419.2211432
  - config_name: FiFA-1k
    features:
      - name: are_different
        dtype: bool
      - name: best_image_uid
        dtype: string
      - name: caption
        dtype: string
      - name: created_at
        dtype: timestamp[ns]
      - name: has_label
        dtype: bool
      - name: image_0_uid
        dtype: string
      - name: image_0_url
        dtype: string
      - name: image_1_uid
        dtype: string
      - name: image_1_url
        dtype: string
      - name: jpg_0
        dtype: binary
      - name: jpg_1
        dtype: binary
      - name: label_0
        dtype: float64
      - name: label_1
        dtype: float64
      - name: model_0
        dtype: string
      - name: model_1
        dtype: string
      - name: ranking_id
        dtype: int64
      - name: user_id
        dtype: int64
      - name: num_example_per_prompt
        dtype: int64
      - name: __index_level_0__
        dtype: int64
    splits:
      - name: train
        num_bytes: 322345306.4634635
        num_examples: 960
    download_size: 318043276
    dataset_size: 322345306.4634635
  - config_name: FiFA-20k
    features:
      - name: are_different
        dtype: bool
      - name: best_image_uid
        dtype: string
      - name: caption
        dtype: string
      - name: created_at
        dtype: timestamp[ns]
      - name: has_label
        dtype: bool
      - name: image_0_uid
        dtype: string
      - name: image_0_url
        dtype: string
      - name: image_1_uid
        dtype: string
      - name: image_1_url
        dtype: string
      - name: jpg_0
        dtype: binary
      - name: jpg_1
        dtype: binary
      - name: label_0
        dtype: float64
      - name: label_1
        dtype: float64
      - name: model_0
        dtype: string
      - name: model_1
        dtype: string
      - name: ranking_id
        dtype: int64
      - name: user_id
        dtype: int64
      - name: num_example_per_prompt
        dtype: int64
      - name: __index_level_0__
        dtype: int64
    splits:
      - name: train
        num_bytes: 6444891471.103872
        num_examples: 19194
    download_size: 6450692192
    dataset_size: 6444891471.103872
  - config_name: FiFA-500
    features:
      - name: are_different
        dtype: bool
      - name: best_image_uid
        dtype: string
      - name: caption
        dtype: string
      - name: created_at
        dtype: timestamp[ns]
      - name: has_label
        dtype: bool
      - name: image_0_uid
        dtype: string
      - name: image_0_url
        dtype: string
      - name: image_1_uid
        dtype: string
      - name: image_1_url
        dtype: string
      - name: jpg_0
        dtype: binary
      - name: jpg_1
        dtype: binary
      - name: label_0
        dtype: float64
      - name: label_1
        dtype: float64
      - name: model_0
        dtype: string
      - name: model_1
        dtype: string
      - name: ranking_id
        dtype: int64
      - name: user_id
        dtype: int64
      - name: num_example_per_prompt
        dtype: int64
      - name: __index_level_0__
        dtype: int64
    splits:
      - name: train
        num_bytes: 162515758.6753295
        num_examples: 484
    download_size: 160479513
    dataset_size: 162515758.6753295
  - config_name: FiFA-50k
    features:
      - name: are_different
        dtype: bool
      - name: best_image_uid
        dtype: string
      - name: caption
        dtype: string
      - name: created_at
        dtype: timestamp[ns]
      - name: has_label
        dtype: bool
      - name: image_0_uid
        dtype: string
      - name: image_0_url
        dtype: string
      - name: image_1_uid
        dtype: string
      - name: image_1_url
        dtype: string
      - name: jpg_0
        dtype: binary
      - name: jpg_1
        dtype: binary
      - name: label_0
        dtype: float64
      - name: label_1
        dtype: float64
      - name: model_0
        dtype: string
      - name: model_1
        dtype: string
      - name: ranking_id
        dtype: int64
      - name: user_id
        dtype: int64
      - name: num_example_per_prompt
        dtype: int64
      - name: __index_level_0__
        dtype: int64
    splits:
      - name: train
        num_bytes: 16102155386.9327
        num_examples: 47955
    download_size: 16114773526
    dataset_size: 16102155386.9327
  - config_name: FiFA-5k
    features:
      - name: are_different
        dtype: bool
      - name: best_image_uid
        dtype: string
      - name: caption
        dtype: string
      - name: created_at
        dtype: timestamp[ns]
      - name: has_label
        dtype: bool
      - name: image_0_uid
        dtype: string
      - name: image_0_url
        dtype: string
      - name: image_1_uid
        dtype: string
      - name: image_1_url
        dtype: string
      - name: jpg_0
        dtype: binary
      - name: jpg_1
        dtype: binary
      - name: label_0
        dtype: float64
      - name: label_1
        dtype: float64
      - name: model_0
        dtype: string
      - name: model_1
        dtype: string
      - name: ranking_id
        dtype: int64
      - name: user_id
        dtype: int64
      - name: num_example_per_prompt
        dtype: int64
      - name: __index_level_0__
        dtype: int64
    splits:
      - name: train
        num_bytes: 1606018334.1820269
        num_examples: 4783
    download_size: 1592623798
    dataset_size: 1606018334.1820269
configs:
  - config_name: FiFA-100k
    data_files:
      - split: train
        path: FiFA-100k/train-*
  - config_name: FiFA-10k
    data_files:
      - split: train
        path: FiFA-10k/train-*
  - config_name: FiFA-1k
    data_files:
      - split: train
        path: FiFA-1k/train-*
  - config_name: FiFA-20k
    data_files:
      - split: train
        path: FiFA-20k/train-*
  - config_name: FiFA-500
    data_files:
      - split: train
        path: FiFA-500/train-*
  - config_name: FiFA-50k
    data_files:
      - split: train
        path: FiFA-50k/train-*
  - config_name: FiFA-5k
    data_files:
      - split: train
        path: FiFA-5k/train-*
    default: true

Pick-a-Pic v2 · FiFA Filtered Subsets

These subsets were produced by filtering the original Pick-a-Pic v2 dataset using FiFA, a data filtering algorithm proposed in the paper Automated Filtering of Human Feedback Data for Aligning Text-to-Image Diffusion Models.

Overview

The filtering process is based on three key metrics:

Preference Margin: Estimated using PickScore
Text Quality: Estimated through LLM scoring
Text Diversity: Estimated using K-NN distance

Dataset Configurations

This dataset provides several configurations, each corresponding to a different number of filtered triplets selected by the automated FiFA algorithm. You can choose from FiFA-500, FiFA-1k, FiFA-5k, FiFA-10k, FiFA-20k, FiFA-50k, and FiFA-100k, depending on your needs. The only difference between these configurations is the number of examples included; all are filtered using the same FiFA method. We recommend using the FiFA-5k configuration as the default, as it generally works best for Stable Diffusion training.

Quick Start

from datasets import load_dataset

# Load a specific configuration
dataset = load_dataset("Dragonjinny/FiFA-pickapic-v2", "FiFA-5k", split="train")

# Access the data
import io
from PIL import Image

for example in dataset:
    caption = example["caption"]  # The prompt text
    jpg_0 = example["jpg_0"]      # First image (bytes)
    jpg_1 = example["jpg_1"]      # Second image (bytes)
    label_0 = example["label_0"]  # Binary label (0 or 1) indicating which image is preferred
    
    # Convert bytes to PIL Images
    image1 = Image.open(io.BytesIO(jpg_0)).convert("RGB")
    image2 = Image.open(io.BytesIO(jpg_1)).convert("RGB")
    
    # Now you can work with the images
    print(f"Caption: {caption}")
    print(f"Preferred image: {'jpg_0' if label_0 == 1 else 'jpg_1'}")
    # image1.show()  # Display the first image
    # image2.show()  # Display the second image

Data Format

Each example contains:

caption: The prompt text
jpg_0: First image (bytes)
jpg_1: Second image (bytes)
label_0: Binary label (0 or 1) indicating which image is preferred
__index_level_0__ : Unique ID for each data point

Citation

If you use this dataset in your research, please cite our paper:

@inproceedings{
yang2025automated,
title={Automated Filtering of Human Feedback Data for Aligning Text-to-Image Diffusion Models},
author={Yongjin Yang and Sihyeon Kim and Hojung Jung and Sangmin Bae and SangMook Kim and Se-Young Yun and Kimin Lee},
booktitle={The Thirteenth International Conference on Learning Representations},
year={2025},
url={https://openreview.net/forum?id=8jvVNPHtVJ}
}

License

This dataset is licensed under MIT License, following the license of the original Pick-a-Pic v2 dataset.