readme fixed.

README.md

@@ -1,47 +1,37 @@
-# XTTS v2 Hugging Face Space (FastAPI)
+---
+title: xtts-v2-api
+emoji: 🗣️
+colorFrom: purple
+colorTo: blue
+sdk: docker
+app_file: app.py
+pinned: false
+---
 
-This folder holds ready-to-push files for a Spaces deployment that exposes XTTS v2 via FastAPI.
+# XTTS v2 FastAPI Space
 
-## Files
-- `app.py` – FastAPI app with `/health` and `/generate` (URL or base64 speaker input).
-- `requirements.txt` – pinned dependencies.
-- `Dockerfile` – runs uvicorn; installs ffmpeg and deps.
-- `spaces.yaml` – tells Spaces to use the Dockerfile.
+FastAPI service exposing XTTS v2 for text-to-speech and voice cloning.
 
-## How to deploy on Hugging Face Spaces
-1) Create a new Space (recommend `Public`, SDK: **Docker**). Name it `xtts-v2-api`.
-2) Clone it locally:
-   ```bash
-   huggingface-cli repo clone spaces/<username>/xtts-v2-api
-   cd xtts-v2-api
-   ```
-3) Copy these files into the Space repo (overwrite if prompted).
-4) Install Git LFS (one time): `git lfs install`.
-5) Commit and push:
-   ```bash
-   git add .
-   git commit -m "Add XTTS v2 FastAPI Space"
-   git push
-   ```
-6) In Space Settings → Secrets, add `SPACE_API_KEY=<your-shared-secret>` (optional but recommended).
-7) In Space Settings → Hardware, choose `T4 GPU` if available. CPU works but slower.
-8) Wait for the build; first startup may take 2–4 minutes. Check Logs if build fails.
+## Endpoints
+- `POST /health` → `{ status: "ok", model: "xtts_v2", device: "cuda|cpu" }`
+- `POST /generate` → returns `audio/wav` bytes
 
-## Smoke tests (after Space is live)
-- Health:
-  ```bash
-  curl -X POST https://<space>.hf.space/health -H "x-api-key: $SPACE_API_KEY"
-  ```
-- Generate (URL speaker):
-  ```bash
-  curl -X POST https://<space>.hf.space/generate \
-    -H "Content-Type: application/json" \
-    -H "x-api-key: $SPACE_API_KEY" \
-    -d '{"text":"Hello from XTTS","speaker_wav":"https://.../sample.wav"}' \
-    --output out.wav
-  ```
+## Auth
+Send header `x-api-key: $SPACE_API_KEY` (set this secret in Space settings).
 
-## Integration notes
-- Next.js should call `/generate` with `text`, `speaker_wav` (signed Supabase URL or base64), optional `language`.
-- Set envs in Vercel: `HF_SPACES_API_URL`, `HF_SPACES_API_KEY`.
-- On errors, the API returns `{ "error": "message" }` with 4xx/5xx.
+## Request body (`/generate`)
+{
+  "text": "Hello from XTTS",
+  "speaker_wav": "https://.../sample.wav or base64 string",
+  "language": "en"
+}## Smoke tests
+curl -X POST https://<space>.hf.space/health \
+  -H "x-api-key: $SPACE_API_KEY"
+
+curl -X POST https://<space>.hf.space/generate \
+  -H "Content-Type: application/json" \
+  -H "x-api-key: $SPACE_API_KEY" \
+  -d '{"text":"Hello from XTTS","speaker_wav":"https://.../sample.wav"}' \
+  --output out.wav## Notes
+- Hardware: request T4 GPU for best latency (CPU works but slower).
+- Env vars for clients: `HF_SPACES_API_URL`, `HF_SPACES_API_KEY`.