| --- |
| title: ZOI PRO |
| emoji: "🧪" |
| colorFrom: blue |
| colorTo: green |
| sdk: gradio |
| sdk_version: 5.49.1 |
| app_file: app.py |
| python_version: "3.11" |
| models: |
| - CarbGeM/ZOI_PRO |
| hf_oauth: true |
| --- |
| |
| # ZOI PRO - HF Image Echo Demo |
|
|
| This Hugging Face Space is a minimal frontend: |
|
|
| - User signs in with Hugging Face (`hf_oauth: true`) |
| - User uploads an image |
| - Frontend sends image + HF token to backend |
| - Backend validates token, checks in-memory quota, and returns the same image |
|
|
| ## Space Secret |
|
|
| - `TRIAL_GATEWAY_URL` (backend base URL, for example `http://localhost:3000`) |
|
|
| ## Backend contract |
|
|
| Frontend calls: |
|
|
| - `POST {TRIAL_GATEWAY_URL}/v1/trial/infer` |
| - Header: `Authorization: Bearer <hf_token>` |
| - Body: `multipart/form-data` with file field `image` |
|
|
| Expected backend behavior: |
|
|
| - Validate HF token against Hugging Face API |
| - Track remaining uses in memory by HF user |
| - Return image bytes unchanged |
| - Return `x-remaining-calls` response header |
|
|
