README.md

metadata

dataset_info:
  features:
    - name: timestamp
      dtype: large_string
    - name: tweet
      dtype: large_string
    - name: language
      dtype: string
    - name: total_tokens
      dtype: int64
    - name: sentiment_class
      list:
        - name: label
          dtype: string
        - name: score
          dtype: float64
    - name: sentiment_label
      dtype: string
    - name: sentiment_score
      dtype: float64
    - name: confidence_level
      dtype: string
    - name: year
      dtype: string
  splits:
    - name: train
      num_bytes: 6772592588.26901
      num_examples: 22033090
  download_size: 3584763878
  dataset_size: 6772592588.26901
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*