DhruvTambekar24/deepseek-ocr-2-demo1
0
๐ DeepSeek-OCR-2 Space Demo
This Hugging Face Space provides an interactive demo for [DeepSeek-OCR-2](https://huggingface.co/deepseek-ai/DeepSeek-OCR-2), a state-of-the-art vision-language model powered by DeepEncoder V2.
๐ Key Features
- Visual Causal Flow: Processes documents in human-like reading order rather than standard top-left to bottom-right scans.
- Markdown Conversion: Preserves document structure, headers, and formatting.
- Table & Form Extraction: Parses complex tabular data and multi-column documents accurately.
- ZeroGPU Accelerated: Uses Hugging Face
@spaces.GPUdynamic allocation for fast, free inference.
๐ Prompt Options
- Document to Markdown (Default):
<image>\n<|grounding|>Convert the document to markdown. - Free OCR:
<image>\nFree OCR. - Detailed Image Description:
<image>\nDescribe this image in detail.
