---
dataset_info:
  features:
  - name: utterance_id
    dtype: large_string
  - name: candidate_index
    dtype: int64
  - name: prediction_text
    dtype: large_string
  - name: token_id_seq
    dtype: large_string
  - name: token_seq
    dtype: large_string
  - name: token_logprob_seq
    dtype: large_string
  splits:
  - name: train
    num_bytes: 225053875
    num_examples: 305325
  download_size: 33248805
  dataset_size: 225053875
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---