README.md

---
language:
- en
license: cc-by-nc-4.0
size_categories:
- 1K<n<10K
task_categories:
- question-answering
- text-generation
- summarization
- text-generation
pretty_name: Situational Instructions Database (SID)
dataset_info:
  features:
  - name: scan
    dtype: string
  - name: scenario
    dtype: string
  - name: objects
    sequence:
    - name: global_id
      dtype: string
    - name: id
      dtype: string
    - name: label
      dtype: string
    - name: ply_color
      dtype: string
    - name: affordances
      sequence: string
    - name: attributes
      struct:
      - name: texture
        sequence: string
      - name: lexical
        sequence: string
      - name: color
        sequence: string
      - name: material
        sequence: string
      - name: shape
        sequence: string
  - name: instructions
    dtype: string
  - name: conversation
    sequence:
    - name: role
      dtype: string
    - name: content
      dtype: string
  splits:
  - name: train
    num_bytes: 32766197
    num_examples: 5519
  - name: test
    num_bytes: 8174412
    num_examples: 1408
  download_size: 12031105
  dataset_size: 40940609
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
paper: arxiv.org/abs/2406.13302
---