[Question]: using minerU(vlm-vllm)

[GXKIM]

Self Checks

• [x] I have searched for existing issues search for existing issues, including closed ones.
• [x] I confirm that I am using English to submit this report (Language Policy).
• [x] Non-english title submitions will be closed directly ( 非英文标题的提交将会被直接关闭 ) (Language Policy).
• [x] Please do not modify this template :) and fill in all the required fields.

Describe your problem

I set ‘vlm-transformers’ in my docker/.env file, but it’s still calling the default pipeline. Here are my logs. ###info ###postman

[dosubot[bot]]

To use MinerU with the vlm-transformers backend, you need to set the environment variable MINERU_BACKEND=vlm-transformers in your docker/.env file (not just 'vlm-transformers' by itself). After making this change, restart all Docker containers for the new setting to take effect. You can verify the active environment inside the container to ensure MINERU_BACKEND is set as expected. If it still defaults to the original pipeline, check for typos, duplicate/conflicting environment variable settings, or custom overrides in your deployment. The backend selection logic is based on MINERU_BACKEND, which supports 'pipeline', 'vlm-transformers', 'vlm-vllm-engine', and 'vlm-http-client' as valid options (docs/faq, mineru_parser.py).

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}

   

[GXKIM]

Wait, that’s not right. vlm-transformers refers to an online model from Hugging Face. Since I’m running a model locally with LM Studio, should I be using vlm-http-client instead? This is assuming it can connect even though the service isn’t deployed via vllm.

env

Enable DocLing

USE_DOCLING=true

Enable Mineru

USE_MINERU=true MINERU_EXECUTABLE="$HOME/uv_tools/.venv/bin/mineru" MINERU_DELETE_OUTPUT=0 # keep output directory MINERU_BACKEND=vlm-http-client # or another backend you prefer MINERU_SERVER_URL=ip:11438

pptx support

DOTNET_SYSTEM_GLOBALIZATION_INVARIANT=1

Even after I made this change, it’s still using the YOLO model and not the multimodal model

[Magicbook1108]

We are working on using mineru as one of the model providers like siliconflow. Soon will be released.

[channlang]

We are working on using mineru as one of the model providers like siliconflow. Soon will be released.

Will it support miner cluster configuration?

[KevinHuSh]

Dose MinerU support cluster mode? Do you have reference about that?

[GXKIM]

###env ‘USE_MINERU=true MINERU_EXECUTABLE="$HOME/uv_tools/.venv/bin/mineru" MINERU_DELETE_OUTPUT=0 # keep output directory MINERU_BACKEND=vlm-http-client # or another backend you prefer MINERU_SERVER_URL=http://10.xxxx:30000’

###miner u server

I’m trying to use a VLM for multimodal parsing, but I’m unable to use the vlm-http-client. It seems I can’t connect to the service.

[yongtenglei]

Hi, @GXKIM

Which version are you using? I just tested it yesterday, it works well.