README.md

metadata

dataset_info:
  features:
    - name: INSTRUCTION
      dtype: string
    - name: RESPONSE
      dtype: string
    - name: SOURCE
      dtype: string
    - name: text
      dtype: string
    - name: question
      dtype: string
    - name: target
      dtype: string
  splits:
    - name: train
      num_bytes: 117622037
      num_examples: 110214
  download_size: 24688336
  dataset_size: 117622037
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: apache-2.0
language:
  - en
pretty_name: A Variety of Combined Logic Data

This is just a concatenation of several other data sets mostly converted to Alpaca style prompts to help give a good logic data set to settle some models or fine tune.