---
dataset_info:
  features:
  - name: title
    dtype: string
  - name: topic
    dtype: string
  - name: text
    dtype: string
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: labels
    sequence: int64
  - name: packed_length
    dtype: int64
  splits:
  - name: train
    num_bytes: 79816412.66475745
    num_examples: 16254
  - name: validation
    num_bytes: 4203448.335242548
    num_examples: 856
  download_size: 28949129
  dataset_size: 84019861.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---