---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: context
    dtype: string
  - name: context_neg
    dtype: string
  splits:
  - name: train
    num_bytes: 22980369
    num_examples: 8000
  - name: validation
    num_bytes: 2830159
    num_examples: 1000
  - name: test
    num_bytes: 2882826
    num_examples: 1000
  download_size: 16233011
  dataset_size: 28693354
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---