Tool-Planner - GitLite

KO | EN

gitlite — search

#hacktoberfest #react #ai #golang #android #machine-learning #rust #linux #docker #nodejs #cli #kubernetes #awesome-list #openai

Tool-Planner

(★ 116)

[ICLR 2025] Tool-Planner: Task Planning with Clusters across Multiple Tools

파일 탐색기

최종 버전 다운로드 (.zip)

- eval.png
- main_result.png
- tool-planner.png
- - GPTFactory.py
- - sim.py
- run.sh
- - plan.py
- - prompt_template.py
- - loader.py
  - preprocessing.py
- tool_library_example.json
- toolkit_library_example.json
- - qa_pipeline.py
  - qa_pipeline_open_domain.py
  - server.py
  - toolbench_server.py
  - utils.py
- - __init__.py
  - apply_delta.py
  - compression.py
  - make_delta.py
  - model_adapter.py
- - api_evaluator.py
  - inference_example.py
  - train.py
- - __init__.py
  - automatic_eval_sample.py
  - convert_answers.py
  - convert_to_answer_format.py
  - eval_and_update_leaderboard.py
  - eval_pass_rate.py
  - eval_preference.py
  - evaluators_comparison.py
  - README.md
  - README_ZH.md
  - requirements.txt
  - run_convert_answer.sh
  - run_pass_rate.sh
  - run_preference.sh
  - utils.py
- - llama_condense_monkey_patch.py
  - llama_flash_attn_monkey_patch.py
  - train.py
  - train_lora.py
  - train_mem.py
- tool_conversation.py
- utils.py
.gitignore
config.py
main.py
README.md
requirement.txt
utils.py

# 설치 가이드

1. 코드 내려받기

git clone https://github.com/OceannTwT/Tool-Planner

깃허브에서 프로젝트 코드 전체를 내 컴퓨터로 내려받습니다.

cd Tool-Planner

방금 내려받은 프로젝트 폴더 안으로 이동합니다.

2. Python

쉬움 추천

사전 준비물

Git GitHub에서 프로젝트 코드를 내려받으려면 필요합니다.
Python 3 설치 시 'Add Python to PATH' 옵션을 꼭 체크하세요 (Windows).

pip install -r requirement.txt

requirements.txt 등에 명시된 파이썬 라이브러리를 설치합니다.

✅ 에러 메시지 없이 실행되고 터미널에 안내 문구가 출력되면 정상입니다.

이 레포의 README에 적힌 실제 명령어를 그대로 가져왔습니다.

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Tool-Planner: Task Planning with Clusters across Multiple Tools

<p align="center">
<img src='https://img.shields.io/badge/Code%20License-Apache_2.0-green.svg'></a>
<img src="https://img.shields.io/badge/python-3.11-blue.svg" alt="Python Version 3.11">
<img src='https://img.shields.io/badge/Data%20License-CC%20By%20NC%204.0-red.svg'>
<a href="https://github.com/OceannTwT/Tool-Planner/stargazers"><img src="https://img.shields.io/github/stars/OceannTwT/Tool-Planner" alt="GitHub Stars"></a>
</p>

This is the official repository of the paper: [Tool-Planner: Dynamic Solution Tree Planning for Large Language Model with Tool Clustering](https://arxiv.org/abs/2406.03807)

![Framework of Tool-Planner.](https://raw.githubusercontent.com/OceannTwT/Tool-Planner/main/assets/tool-planner.png)

## 🎯Installation Enviorment

Update your environment for the required dependency.

```shell
conda create --name tool-planner python=3.11 -y
conda activate tool-planner
pip install -r requirement.txt
```

## Collect ToolBench Util

Use the following link to download the `Toolbench` dataset. It will be used to extract the tool calling methods and descriptions from `RapidAPI`, and to complete the subsequent setup of the Toolkit.

[Google Drive](https://drive.google.com/drive/folders/1TysbSWYpP8EioFu9xPJtpbJZMLLmwAmL) or [Tsinghua Cloud](https://cloud.tsinghua.edu.cn/f/c9e50625743b40bfbe10/)

* Putting the file to the main directory.

```shell
mv your_path/datas Tool-Planner
```

* Make sure you have the access of `ToolBench` API. refer to [Toolbench](https://github.com/OpenBMB/ToolBench) to apply for the permission.

## Download SimCSE

Downloading SimCSE model to your local directory.

```shell
mv your_path/sup-simcse-roberta-base Tool-Planner/model
```

Or Upload it from `HuggingFace`.

```python
tokenizer = AutoTokenizer.from_pretrained("princeton-nlp/sup-simcse-roberta-base")
model = AutoModel.from_pretrained("princeton-nlp/sup-simcse-roberta-base")
```

## Inference Tool-Planner

Use your `ChatGPT` API and `Toolbench` API, replace the API key in `run.sh`, and run the following script.

```shell
bash script/run.sh
```

To obtain the corresponding result information for your query, change the number of toolkits and adjust the file location of the input query.

```shell
export TOOLBENCH_KEY=""
export OPENAI_KEY=""
export PYTHONPATH=./

python main.py \
    --toolkit_dir src/toolkits \
    --tool_api_dir datas/toolenv/tools/ \
    --backbone_model gpt_3.5 \
    --toolkit_num 20 \
    --openai_key $OPENAI_KEY \
    --tool_env datas/toolenv/tools/ \
    --tool_output_file tool_lib/my_tool_library.json \
    --toolkit_output_file tool_lib/my_toolkit_library.json \
    --input_query_file data/instruction/G3_query.json \
    --output_answer_file data/instruction/tool_result.json \
    --simcse_file model_lib/sup-simcse-roberta-base \
    --toolbench_key $TOOLBENCH_KEY
```

In the tool_lib, we provide a example of the toolkit, corresponding to the generated toolkit contents, change your output tool and toolkit file on `--toolkit_output_file` and `--tool_output_file`.

## Main Results

![Main Result of Tool-Planner](https://raw.githubusercontent.com/OceannTwT/Tool-Planner/main/assets/main_result.png)

## Evaluations

![Time and Performance Evaluation](https://raw.githubusercontent.com/OceannTwT/Tool-Planner/main/assets/eval.png)

- Experiments have demonstrated that our method exhibits competitive performance and efficient performance configuration under different approaches.

## 🛠️Update logs

[24/01/23] 🔥 Our paper was accepted on ICLR 2025!
[24/06/09] 🔥 We have released the version 1.0.0 for Tool-Planner.

## 🔗Citation

If you use this codebase or Tool-Planner inspires your work, we would greatly appreciate it if you could star the repository and cite it using the following BibTeX entry:

```bibtex

@inproceedings{
liu2025toolplanner,
title={Tool-Planner: Task Planning with Clusters across Multiple Tools},
author={Yanming Liu and Xinyue Peng and Jiannan Cao and Shi Bo and Yuwei Zhang and Xuhong Zhang and Sheng Cheng and Xun Wang and Jianwei Yin and Tianyu Du},
booktitle={The Thirteenth International Conference on Learning Representations},
year={2025},
url={https://openreview.net/forum?id=dRz3cizftU}
}

```

## Acknowledgement

We thank the ZJU TRAIL Lab assistance for extending Tool-Planner!