Spaces:
Running on L40S
Running on L40S
| title: Penguin VL 8B | |
| emoji: 💬 | |
| colorFrom: yellow | |
| colorTo: purple | |
| sdk: gradio | |
| sdk_version: "6.5.1" | |
| python_version: "3.11.13" | |
| app_file: app.py | |
| pinned: false | |
| hf_oauth: true | |
| hf_oauth_scopes: | |
| - inference-api | |
| license: apache-2.0 | |
| short_description: "a compact Vision-Language Model" | |
| startup_duration_timeout: 1h | |
| preload_from_hub: | |
| - tencent/Penguin-VL-8B | |
| This Space runs Penguin-VL-8B and preloads model artifacts from `tencent/Penguin-VL-8B`. | |
| On dedicated GPU hardware, the app also loads the model into GPU memory during startup so users can interact without waiting for the first request to initialize the model. | |
| If you move this Space back to ZeroGPU later, restore the `@spaces.GPU(...)` path and set `PRELOAD_MODEL_ON_STARTUP=0` in the Space variables. | |