README.md

metadata

license: apache-2.0
size_categories:
  - 100K<n<1M
pretty_name: evalhrf
configs:
  - config_name: ddpm-celebahq
    data_files:
      - split: train
        path: data/ddpm-celebahq/samples/part-*
dataset_info:
  - config_name: ddpm-celebahq
    features:
      - name: images
        dtype: image
      - name: rewards
        list:
          - name: aesthetic_score
            dtype: float
          - name: file_size_kb
            dtype: float

Dataset Card for Eval Finetuning Diffusion Models with Reinforcement Learning

XYZ