https://github.com/amazon-science/auto-cot

Official implementation for "Automatic Chain of Thought Prompting in Large Language Models" (stay tuned & more will be updated)
https://github.com/amazon-science/auto-cot

chain-of-thought gpt-3 gpt3-prompts gpt3-resources large-language-models prompt-engineering reasoning

Last synced: 3 days ago
JSON representation

Official implementation for "Automatic Chain of Thought Prompting in Large Language Models" (stay tuned & more will be updated)

• Host: GitHub
• URL: https://github.com/amazon-science/auto-cot
• Owner: amazon-science
• License: apache-2.0
• Created: 2022-10-05T07:47:17.000Z (over 2 years ago)
• Default Branch: main
• Last Pushed: 2024-03-13T12:25:02.000Z (10 months ago)
• Last Synced: 2025-01-04T20:05:20.626Z (11 days ago)
• Topics: chain-of-thought, gpt-3, gpt3-prompts, gpt3-resources, large-language-models, prompt-engineering, reasoning
• Language: Jupyter Notebook
• Homepage: https://arxiv.org/abs/2210.03493
• Size: 56.6 KB
• Stars: 1,636
• Watchers: 17
• Forks: 147
• Open Issues: 9
• Metadata Files:
  • Readme: README.md
  • Contributing: CONTRIBUTING.md
  • License: LICENSE
  • Code of conduct: CODE_OF_CONDUCT.md

Awesome Lists containing this project

• awesome-instruction-datasets - Auto CoT - takeshi188/zero_shot_cot/dataset](https://github.com/kojima-takeshi188/zero_shot_cot/tree/main/dataset) \|[kojima-takeshi188/zero_shot_cot/log](https://github.com/kojima-takeshi188/zero_shot_cot/tree/main/log) | amazon-science | | EN | | | | | [download](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT/tree/main/Auto-CoT) | (Statistics)
• StarryDivineSky - amazon-science/auto-cot - CoT 使用更多的cheers和多样性来节省思维链提示设计中的巨大手动工作，匹配甚至超过 GPT-3 上的手动设计性能。 (A01_文本生成_文本对话 / 大语言对话模型及数据)

README

        
# Auto-CoT: Automatic Chain of Thought Prompting in Large Language Models (ICLR 2023)

[![Open Auto-CoT in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/amazon-science/auto-cot/blob/main/try_cot_colab.ipynb)

Cheer AI up with the "let's think step by step" prompt? More plz. *Let’s think not just step by step, but also one by one.*

Auto-CoT uses more cheers & diversity to SAVE huge manual efforts in chain of thought prompt design, matching or even exceeding performance of manual design on GPT-3.

Check out our [25-page paper](https://arxiv.org/pdf/2210.03493.pdf) for more information.

![](https://user-images.githubusercontent.com/22279212/194787183-a1f8dff8-a0ad-43a1-827f-819671503860.png)

![](https://user-images.githubusercontent.com/22279212/194787130-d28c9191-588c-41d2-a259-62377f19c934.png)

## Requirements

Python>=3.8

```

pip install torch==1.8.2+cu111 torchtext==0.9.2 -f https://download.pytorch.org/whl/lts/1.8/torch_lts.html

pip install -r requirements.txt

```

## Datasets

Download the datasets from the following:

```

https://github.com/kojima-takeshi188/zero_shot_cot/tree/main/dataset

https://github.com/kojima-takeshi188/zero_shot_cot/tree/main/log

```

## Quick Start

See ```try_cot.ipynb```

## Instructions

Construct Demos:

```

python run_demo.py \

--task multiarith \

--pred_file log/multiarith_zero_shot_cot.log \

--demo_save_dir demos/multiarith

```

Run inference:

```

python run_inference.py \

--dataset multiarith \

--demo_path demos/multiarith \

--output_dir experiment/multiarith

```

## Citing Auto-CoT

```

@inproceedings{zhang2023automatic,

  title={Automatic Chain of Thought Prompting in Large Language Models},

  author={Zhang, Zhuosheng and Zhang, Aston and Li, Mu and Smola, Alex},

  booktitle={The Eleventh International Conference on Learning Representations (ICLR 2023)},

  year={2023}

}

```

## Security

See [CONTRIBUTING](CONTRIBUTING.md#security-issue-notifications) for more information.

## License

This project is licensed under the Apache-2.0 License.