README.md

---
dataset_info:
  features:
  - name: idx
    dtype: string
  - name: model
    dtype: string
  - name: dataset
    dtype: string
  - name: task_id
    dtype: string
  - name: meta_json
    dtype: string
  - name: model_output
    dtype: string
  - name: closest_gt
    dtype: string
  - name: hallucination_token_index
    dtype: int64
  - name: tokens_json
    dtype: string
  - name: token_types_json
    dtype: string
  - name: token_logprobs_json
    dtype: string
  - name: execution_json
    dtype: string
  - name: claim
    dtype: string
  splits:
  - name: train
    num_bytes: 49209234409
    num_examples: 81626
  download_size: 4971229879
  dataset_size: 49209234409
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---