README.md

---
dataset_info:
  features:
  - name: collection_slug
    dtype: string
  - name: item_1
    struct:
    - name: caption_finish_reasons
      sequence: string
    - name: caption_logprobs
      sequence: float64
    - name: caption_model_name
      sequence: string
    - name: caption_num_tokens
      sequence: int64
    - name: caption_temperature
      sequence: float64
    - name: caption_texts
      sequence: string
    - name: collection_id
      dtype: int64
    - name: collection_slug
      dtype: string
    - name: item_id
      dtype: int64
    - name: item_license
      dtype: string
    - name: item_license_link
      dtype: string
    - name: item_license_owner
      dtype: string
    - name: item_slug
      dtype: string
    - name: item_svg
      dtype: string
    - name: item_tags
      sequence: string
    - name: item_title
      dtype: string
  - name: item_2
    struct:
    - name: caption_finish_reasons
      sequence: string
    - name: caption_logprobs
      sequence: float64
    - name: caption_model_name
      sequence: string
    - name: caption_num_tokens
      sequence: int64
    - name: caption_temperature
      sequence: float64
    - name: caption_texts
      sequence: string
    - name: collection_id
      dtype: int64
    - name: collection_slug
      dtype: string
    - name: item_id
      dtype: int64
    - name: item_license
      dtype: string
    - name: item_license_link
      dtype: string
    - name: item_license_owner
      dtype: string
    - name: item_slug
      dtype: string
    - name: item_svg
      dtype: string
    - name: item_tags
      sequence: string
    - name: item_title
      dtype: string
  - name: similarity
    dtype: float64
  - name: instruction
    dtype: string
  splits:
  - name: train
    num_bytes: 2281819762
    num_examples: 269106
  - name: validation
    num_bytes: 1214438
    num_examples: 200
  - name: test
    num_bytes: 12718293
    num_examples: 2000
  download_size: 274954282
  dataset_size: 2295752493
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---