🎉Upgrade Notice MinerU is a LLM-powered tool that converts PDFs into machine-readable formats. 0.7.1 is now available, which add new integration option of the paddle tablemaster table recognition model, enhancing table processing speed. 🚀 github.com/opendatalab/Mi…

We present DocLayout-YOLO, which is suitable for diverse document layout detection, including but not limited to papers, textbooks, test papers, slides and other document types. ✨github:github.com/opendatalab/Do… 📜paper:arxiv.org/abs/2410.12628 💻demo:huggingface.co/spaces/opendat…

📢MinerU New Year Update – January 2025 Highlights There are the new features: Brand Visual Revamp: A complete redesign of the MinerU brand, along with the official relaunch of our website, offering easy access to technical documentation. Visit us at: mineru.net

● Official Client Release: Download and use with no programming required. Simply drag and drop to quickly process multiple documents for extraction without the need for login.

● Online API Services & Demo: Aligned with the latest model capabilities, optimized resource scheduling strategies, and enhanced batch processing capabilities.

‼️ Important Notice 📷: The v2 and v3 version APIs are now discontinued. Please migrate to the new v4 API, available under the new domain, and create a new token for continued use.

"Wanjuan2.0" is a multilingual and multimodal #corpus that comprises four #data modalities: full text, image- text, video, and #SFT, totaling 11.5 million data entries, covering Russian, Arabic, Korean, Vietnamese, Thai, etc. Open-sourse link: opendatalab.com/applyMultiling…

MinerU's core function at a glance:

The multilingual and multimodal #corpus "Wanjuan2.0" was open-sourced on HuggingFace, with ultra-fine #data , and applicable to multiple scenarios, such as cultural tourism, commercial trade, science and technology education. FREE DOWNLOAD FROM: huggingface.co/datasets?sort=…

What is your ideal data processing #tool? Get #MinerU as a professional assistant to help you get #AI-READY #data . Find out the core function of MinerU as your wish!

Are you looking for a tool to help you labeling #data? You can try #LabelU, the flexible labeling tool, which is applicable to #CV, voice interaction and #AI-assisted labeling. 👉labelu.shlab.tech/tasks/

Document content analysis has been a crucial research area in computer vision. We present #MinerU, an open-source solution for high-precision document content extraction. Deep dive into MinerU via the technical report: mineru.site/Saaas%E6%9C%8D…

We are very pleased to know that one of our users just launched a website about #MinerU! The website has deployed open-source solutions for data processing, tutoring, sharing of usage experience, etc. Welcome to join the community : mineru.site

The open-source dataset WanJuanSiLu, designed to provide high-quality training corpora for low-resource languages, thereby advancing the research and development of multilingual models. WanJuanSiLu mainly consists of eight subsets: Thai, Russian, Arabic, Korean, Hungarian, etc.

Agentic Document Extraction just got much faster! From previous 135sec median processing time down to 8sec. Extracts not just text but diagrams, charts, and form fields from PDFs to give LLM-ready output. Please see the video for details and some application ideas.