metadata

size_categories: n<1K
dataset_info:
  features:
    - name: prompt
      dtype: string
    - name: models
      sequence: string
    - name: images
      list:
        - name: path
          dtype: string
  splits:
    - name: train
      num_bytes: 615
      num_examples: 2
  download_size: 3357
  dataset_size: 615
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
tags:
  - synthetic
  - distilabel
  - rlaif

Dataset Card for img-prefs-distilabel-artifacts-sample

This dataset has been created with distilabel.

Dataset Summary

This dataset contains a pipeline.yaml which can be used to reproduce the pipeline that generated it in distilabel using the distilabel CLI:

distilabel pipeline run --config "https://huggingface.co/datasets/dvilasuero/img-prefs-distilabel-artifacts-sample/raw/main/pipeline.yaml"

or explore the configuration:

distilabel pipeline info --config "https://huggingface.co/datasets/dvilasuero/img-prefs-distilabel-artifacts-sample/raw/main/pipeline.yaml"

Dataset structure

The examples have the following structure per configuration:

Configuration: default

{
    "images": [
        {
            "path": "artifacts/flux_schnell/images/90b884933d23c4d57ca01dbe2898d405.jpeg"
        },
        {
            "path": "artifacts/flux_dev/images/90b884933d23c4d57ca01dbe2898d405.jpeg"
        }
    ],
    "models": [
        "black-forest-labs/FLUX.1-schnell",
        "black-forest-labs/FLUX.1-dev"
    ],
    "prompt": "intelligence"
}

This subset can be loaded as:

from datasets import load_dataset

ds = load_dataset("dvilasuero/img-prefs-distilabel-artifacts-sample", "default")

Or simply as it follows, since there's only one configuration and is named default:

from datasets import load_dataset

ds = load_dataset("dvilasuero/img-prefs-distilabel-artifacts-sample")

Artifacts

Step: flux_dev
- Artifact name: images
  - type: image
  - library: diffusers
Step: flux_schnell
- Artifact name: images
  - type: image
  - library: diffusers