---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: input
    dtype: string
  - name: output
    dtype: string
  - name: id
    dtype: float64
  splits:
  - name: train
    num_bytes: 2487281447
    num_examples: 3229975
  download_size: 1513934252
  dataset_size: 2487281447
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: apache-2.0
task_categories:
- text-generation
- question-answering
language:
- en
- zh
tags:
- cot
- conversational
pretty_name: cot reformatted
size_categories:
- 1M<n<10M
---
# Dataset Card for "CoT_reformatted"

This dataset is reformatted from: QingyiSi/Alpaca-CoT

All credit goes there. Thanks to QingyiSi for the work in consolidating many diverse sources for comparison and cross-file analysis.

There were some issues loading files from that dataset for a testing project. 

I extracted the following data files for this subset:

- alpaca_data_cleaned
- CoT_data
- firefly       
- instruct
- alpaca_gpt4_data
- dolly 
- GPTeacher
- thoughtsource
- finance_en
- instinwild_en