Spaces:
Running on Zero
Running on Zero
| title: Doc-Split Demo | |
| emoji: 📄✂️ | |
| colorFrom: indigo | |
| colorTo: blue | |
| sdk: gradio | |
| sdk_version: 6.20.0 | |
| app_file: app.py | |
| short_description: Split a merged PDF into its constituent documents | |
| pinned: false | |
| # Doc-Split Demo | |
| Upload a PDF that concatenates several documents; the model marks where each new document begins and groups | |
| the pages into documents. Pick **doc-split-v2** (commercial flagship) or the open **doc-split-v1**. Uses the PDF's embedded | |
| text layer where present; scanned pages fall back to the image model. Confidence is calibrated. | |
| Part of the doc-split hub: [leaderboard](https://huggingface.co/spaces/nutrientdocs/doc-split-leaderboard) · | |
| [flagship](https://huggingface.co/nutrientdocs/doc-split-v2) · [open v1](https://huggingface.co/nutrientdocs/doc-split-v1) · | |
| [benchmark](https://huggingface.co/datasets/nutrientdocs/doc-split-benchmark). | |
| ## Secret required | |
| **doc-split-v2** is commercial and loaded from a **private** repo, so this Space needs an `HF_TOKEN` | |
| **secret** (Settings → Variables and secrets) with read access to it. The weights are used server-side only | |
| and are never downloadable. Without the secret, the v2 model is marked _unavailable_ and the open **doc-split-v1** | |
| still runs. Inference runs on **ZeroGPU**. | |