---
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: image
    dtype: image
  - name: caption
    dtype: string
  - name: caption_ls
    list: string
  - name: category
    dtype: string
  - name: width
    dtype: int32
  - name: height
    dtype: int32
  - name: image_phash
    dtype: string
  - name: text_length
    dtype: int32
  - name: word_count
    dtype: int32
  - name: num_tokens_bert
    dtype: int32
  - name: num_tokens_gpt
    dtype: int32
  - name: num_faces
    dtype: int32
  - name: clip_similarity_vitb32
    dtype: float32
  - name: clip_similarity_vitl14
    dtype: float32
  - name: nsfw_score_opennsfw2
    dtype: float32
  - name: nsfw_score_gantman
    dtype: float32
  - name: watermark_score
    dtype: float32
  - name: aesthetic_score_laion_v2
    dtype: float32
  splits:
  - name: train
    num_bytes: 1040217707985.875
    num_examples: 14552129
  download_size: 1026963871833
  dataset_size: 1040217707985.875
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---
이유는 모르겠는데, height와 width가 똑같은 데이터만 남았음.
일단 다운은 끝났으니 사용할 수 도 있으니 올려둠