---
license: mit
dataset_info:
  features:
  - name: created_at
    dtype: timestamp[ns]
  - name: tweet
    dtype: string
  - name: frames
    sequence: string
  splits:
  - name: train
    num_bytes: 814652227
    num_examples: 3111140
  download_size: 465399524
  dataset_size: 814652227
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---