---
dataset_info:
  features:
  - name: task_id
    dtype: int64
  - name: reward
    dtype: float64
  - name: truncated_messages
    dtype: string
  - name: removed_messages
    dtype: string
  - name: compressed_convo
    dtype: string
  - name: original_conversation_length
    dtype: int64
  - name: truncated_conversation_length
    dtype: int64
  splits:
  - name: train
    num_bytes: 238192
    num_examples: 10
  download_size: 61001
  dataset_size: 238192
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---