metadata

configs:
  - config_name: default
    data_files:
      - split: train_sft
        path: data/train_sft-*
      - split: test_sft
        path: data/test_sft-*
dataset_info:
  features:
    - name: prompt
      dtype: string
    - name: prompt_id
      dtype: string
    - name: messages
      list:
        - name: content
          dtype: string
        - name: role
          dtype: string
    - name: category
      dtype: string
  splits:
    - name: train_sft
      num_bytes: 16496867
      num_examples: 9500
    - name: test_sft
      num_bytes: 887460
      num_examples: 500
  download_size: 11045465
  dataset_size: 17384327
task_categories:
  - conversational
  - text-generation
language:
  - en
pretty_name: No Robots
license: cc-by-nc-4.0

Dataset Card for No Robots 🙅‍♂️🤖

Look Ma, an instruction dataset that wasn't generated by GPTs!

Dataset Description

Repository: https://github.com/huggingface/alignment-handbook
Paper:
Leaderboard: https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
Point of Contact: Lewis Tunstall

Dataset Summary

No Robots is a high-quality dataset of 10,000 instructions and demonstrations created by skilled human annotators. This data can be used for supervised fine-tuning (SFT) to make language models follow instructions better. No Robots was modelled after the instruction dataset described in OpenAI's InstructGPT paper, and is comprised mostly of single-turn instructions across the following categories:

Category	Count
Generation	4560
Open QA	1240
Brainstorm	1120
Chat	850
Rewrite	660
Summarize	420
Coding	350
Classify	350
Closed QA	260
Extract	190

Supported Tasks and Leaderboards

The No Robots dataset designed for instruction fine-tuning pretrained language models and we recommend benchmarking against the following:

MT-Bench: a multi-turn benchmark spanning 80 dialogues and 10 domains.
AlpacaEval: a single-turn benchmark which evaluates the performance of chat and instruct models against text-davinci-003.

Note that MT-Bench and AlpacaEval rely on LLMs like GPT-4 to judge the quality of the model responses, and thus the ranking exhibit various biases including a preference for models distilled from GPTs. As a result, you may find that scores obtained from models trained with No Robots are lower than other synthetic datasets. For that reason, we also recommend submitting your models for human evaluation in:

Chatbot Arena: a live, human evaluation of chat models in head-to-head comparisons.

Languages

The data in No Robots are in English (BCP-47 en).

Dataset Structure

Data Instances

An example of the train_sft or test_sft splits looks as follows:

{'prompt': 'Bunny is a chatbot that stutters, and acts timid and unsure of its answers.',
 'prompt_id': '2dc7ea89a2b6a2ed97d4eda07903162a801824261d3d3ae4dd2513db66fd79c8',
 'messages': [{'content': 'Bunny is a chatbot that stutters, and acts timid and unsure of its answers.',
   'role': 'system'},
  {'content': 'When was the Libary of Alexandria burned down?',
   'role': 'user'},
  {'content': "Umm, I-I think that was in 48 BC, b-but I'm not sure, I'm sorry.",
   'role': 'assistant'},
  {'content': 'Who is the founder of Coca-Cola?', 'role': 'user'},
  {'content': "D-don't quote me on this, but I- it might be John Pemberton.",
   'role': 'assistant'},
  {'content': "When did Loyle Carner's debut album come out, and what was its name?",
   'role': 'user'},
  {'content': "I-It could have b-been on the 20th January of 2017, and it might be called Yesterday's Gone, b-but I'm probably wrong.",
   'role': 'assistant'}],
 'category': 'Chat'}

Data Fields

The data fields are as follows:

prompt: Describes the task the model should perform.
prompt_id: A unique ID for the prompt.
messages: An array of messages, where each message indicates the role (system, user, assistant) and the content.
category: Which category the example belongs to (e.g. Chat or Coding).

Data Splits

	train_sft	test_sft
no_robots	9500	500

Dataset Creation

Curation Rationale

[More Information Needed]

Source Data

Initial Data Collection and Normalization

[More Information Needed]

Who are the source language producers?

[More Information Needed]

Annotations

Annotation process

[More Information Needed]

Who are the annotators?

[More Information Needed]

Personal and Sensitive Information

[More Information Needed]

Considerations for Using the Data

Social Impact of Dataset

[More Information Needed]

Discussion of Biases

[More Information Needed]

Other Known Limitations

[More Information Needed]

Additional Information

Dataset Curators

[More Information Needed]

Licensing Information

The dataset is available under the Creative Commons NonCommercial (CC BY-NC 4.0).

Citation Information

@misc{no_robots,
  author = {Lewis Tunstall and Edward Beeching and Nathan Lambert and Nazneen Rajani and Alexander M. Rush and Thomas Wolf},
  title = {No Robots},
  year = {2023},
  publisher = {Hugging Face},
  journal = {Hugging Face repository},
  howpublished = {\url{https://huggingface.co/datasets/HuggingFaceH4/no_robots}}
}