---
license: apache-2.0
task_categories:
- text-generation
language:
- en
tags:
- language-agent
- web-agent
- web-browsing
- reasoning
- planning
size_categories:
- 1K<n<10K
---

# 🪄 Agent Lumos: Unified and Modular Training for Open-Source Language Agents
<p align="center">
  🌐<a href="https://allenai.github.io/lumos">[Website]</a> &nbsp;
  📝<a href="https://arxiv.org/abs/2311.05657">[Paper]</a> &nbsp;
  🤗<a href="https://huggingface.co/datasets?sort=trending&search=ai2lumos">[Data]</a> &nbsp;
  🤗<a href="https://huggingface.co/models?sort=trending&search=ai2lumos">[Model]</a> &nbsp;
  🤗<a href="https://huggingface.co/spaces/ai2lumos/lumos_data_demo">[Demo]</a> &nbsp;
</p>

We introduce 🪄**Lumos**, Language Agents with **Unified** Formats, **Modular** Design, and **Open-Source** LLMs. **Lumos** unifies a suite of complex interactive tasks and achieves competitive performance with GPT-4/3.5-based and larger open-source agents. 

**Lumos** has following features:
* 🧩 **Modular Architecture**:
  - 🧩 **Lumos** consists of planning, grounding, and execution modules built based on LLAMA-2-7B/13B and off-the-shelf APIs.
  - 🤗 **Lumos** utilizes a unified data format that encompasses multiple task types, thereby enabling the developed agent framework to conveniently support a range of interactive tasks.
* 🌍 **Diverse Training Data**:
  - 🌍 **Lumos** is trained with ~56K diverse high-quality subgoal/action annotations from ground-truth reasoning steps in existing benchmarks with GPT-4.
  - ⚒️ **Lumos** data can be instrumental for future research in developing open-source agents for complex interactive tasks.
* 🚀 **Competitive Performance**:
  - 🚀 **Lumos** is comparable or even beats **GPT-series** agents on web/complex QA tasks Mind2Web and HotpotQA, and **larger open agents** on math and multimodal tasks.
  - 🚀 **Lumos** exceeds contemporaneous agents that have been **fine-tuned** with in-domain HotpotQA, Mind2Web and ScienceQA annotations, such as **FiReAct**, **AgentLM**, and **AutoAct**.
  - 🚀 **Lumos** performs better than open agent baseline formulations including **chain-of-thoughts** and **integrated** training.
  - 🚀 **Lumos** surpasses larger open LLM agents and domain-specific agents on unseen tasks, WebShop and InterCode_SQL.

## Data Overview
`lumos_web_agent_plan_iterative` is the data for training **planning** module on **web agent** task in **Lumos-Iterative (Lumos-I)** formulation. 

The source of the training annotation training data is shown below:

| Task | Number |
|---|---|
|Mind2Web|1009|

## Models Trained with the Data
`lumos_web_agent_plan_iterative` is used to train the following models.

|Model|Huggingface Repo|
|---|---|
|`lumos_web_agent_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_web_agent_plan_iterative) |
|`lumos_web_agent_ground_iterative-13B`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_web_agent_plan_iterative-13B) |
|`lumos_unified_ground_iterative`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_plan_iterative) |
|`lumos_unified_ground_iterative-13B`| [🤗Huggingface Repo](https://huggingface.co/ai2lumos/lumos_unified_plan_iterative-13B) |

## Citation

If you find this work is relevant with your research, please feel free to cite our work!
```
@article{yin2023lumos,
  title={Agent Lumos: Unified and Modular Training for Open-Source Language Agents},
  author={Yin, Da and Brahman, Faeze and Ravichander, Abhilasha and Chandu, Khyathi and Chang, Kai-Wei and Choi, Yejin and Lin, Bill Yuchen},
  journal={arXiv preprint arXiv:2311.05657},
  year={2023}
}
```