| title: Simam3D ZeroGPU | |
| emoji: 🧭 | |
| colorFrom: indigo | |
| colorTo: blue | |
| sdk: gradio | |
| app_file: app.py | |
| suggested_hardware: zero-a10g | |
| pinned: false | |
| models: | |
| - depth-anything/Depth-Anything-V2-Small-hf | |
| - depth-anything/DA3-LARGE | |
| # Simam3D ZeroGPU | |
| Gradio-SDK deployment variant of Simam3D for Hugging Face ZeroGPU. It reuses | |
| the same tested depth, fusion, Gaussian export, provenance, and failure-report | |
| code as the Docker Space. ZeroGPU availability and queue time are platform | |
| controlled; this variant does not claim a successful model run until a real | |
| request returns a validated manifest. | |
| Start with the explicit export smoke test to separate UI/GPU scheduling from | |
| model inference. DA3 remains optional and V2 is the stable baseline. | |