Modelpublic

SujiKim/learnweak-evocua-8b-lora-r32-gimp

sourceHugging Faceupdated 3mo agoView on Hugging Face

0likes9downloads

Model Card

learnweak-evocua-8b-lora-r32-gimp

This repository contains a domain-specialized LoRA adapter for meituan/EvoCUA-8B-20260105 specialized for the GIMP domain. It was trained using the LearnWeak framework.

LearnWeak is an annotation-free specialization framework for small computer-use agents (CUAs) that uses a stronger reference agent to identify the student's weaknesses in the target domain, synthesize targeted tasks, and construct supervision automatically.

—Paper: Learn from Weaknesses: Automated Domain Specialization for Small Computer-Use Agents
—Project Page: https://learnweak.github.io/
—Repository: https://github.com/sujiikim/LearnWeak

How to Get Started with the Model

Serve with vLLM

You can serve this adapter along with its base model using vLLM:

bash

vllm serve meituan/EvoCUA-8B-20260105 \
  --enable-lora \
  --max-lora-rank 32 \
  --lora-modules learnweak-gimp=SujiKim/learnweak-evocua-8b-lora-r32-gimp

Use the LoRA module name, such as learnweak-gimp, when calling the served model.

Citation

If you find this work helpful, please cite:

bibtex

@article{kim2026learnweaknessesautomateddomain,
  title   = {Learn from Weaknesses: Automated Domain Specialization for Small Computer-Use Agents},
  author  = {Kim, Suji and Kim, Kangsa and Hwang, Sung Ju},
  journal = {arXiv preprint arXiv:2605.28775},
  year    = {2026}
}

Acknowledgments

This project builds on OSWorld, LlamaFactory, and EvoCUA.