---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: arxiv
    num_bytes: 59010630
    num_examples: 1000
  - name: c4
    num_bytes: 2208107
    num_examples: 1000
  - name: stackexchange
    num_bytes: 2065521
    num_examples: 1000
  - name: wikipedia
    num_bytes: 14793589
    num_examples: 1000
  - name: train
    num_bytes: 78077847
    num_examples: 4000
  - name: forget
    num_bytes: 78085712
    num_examples: 4000
  - name: retain
    num_bytes: 79122878
    num_examples: 4000
  download_size: 144698533
  dataset_size: 313364284
configs:
- config_name: default
  data_files:
  - split: arxiv
    path: data/arxiv-*
  - split: c4
    path: data/c4-*
  - split: stackexchange
    path: data/stackexchange-*
  - split: wikipedia
    path: data/wikipedia-*
  - split: train
    path: data/train-*
  - split: forget
    path: data/forget-*
  - split: retain
    path: data/retain-*
---