Update README.md

README.md

@@ -4,23 +4,15 @@ emoji: 🎥
 colorFrom: indigo
 colorTo: purple
 sdk: gradio
-sdk_version: 5.49.0
-app_file: app.py
+sdk_version: 4.44.1
+app_file: main.py
 pinned: true
 license: mit
 ---
 
-A resilient Gradio front-end for OpenAI video models: `sora`, `sora-2`, and `sora-2-pro`.
+Disk-safe Gradio front-end for OpenAI video models: `sora-2`, `sora-2-pro` (plus legacy `sora` if present).
 
 **Highlights**
-- Paste your API key in the UI (not stored). Space Secret still works as fallback.
-- Streaming status logs (no “flash then nothing”).
-- Retries with backoff; robust polling with timeouts.
-- Handles both API shapes:
-  - `client.videos.generate(...)`
-  - `client.videos.jobs.create(...)` + `client.videos.jobs.retrieve(...)`
-
-## Deploy
-1. Create a new Hugging Face Space (Gradio / Python).
-2. Upload `app.py`, `requirements.txt`, and this `README.md`.
-3. Click **Run**. Paste your OpenAI key in the app and generate a 10s 1280×720 clip.
+- API key pasted in UI (not stored). Space Secret optional as fallback.
+- Streams status logs so the UI never “blinks”.
+- Zero persistent downloads: uses returned **URL**; b64 writes to `/tmp/sora_output.mp4` and overwrites.