Upload README.md with huggingface_hub

README.md

@@ -1,13 +1,27 @@
 ---
-title: Ltx 2.3 Inpaint
-emoji: 🚀
-colorFrom: purple
-colorTo: purple
+title: LTX-2.3 Video Inpainting
+emoji: 🪄
+colorFrom: red
+colorTo: indigo
 sdk: gradio
-sdk_version: 6.18.0
-python_version: '3.12'
+sdk_version: 6.13.0
+python_version: "3.12"
 app_file: app.py
 pinned: false
+hardware: zero-a10g
+short_description: Mask and regenerate a video region with LTX-2.3
+models:
+  - diffusers/LTX-2.3-Distilled-Diffusers
+  - linoyts/ltx2.3-inpainting-lora
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# 🪄 LTX-2.3 Video Inpainting
+
+Mask a region of a video and regenerate it from a text prompt, keeping the rest of the frame
+intact. Brush over the area to replace on the first frame; the mask is applied across all frames,
+the masked region is fed to the model as the in-context reference, and the output is composited
+back over the original (feathered) so only the masked area changes.
+
+Runs the inpainting IC-LoRA from [`linoyts/ltx2.3-inpainting-lora`](https://huggingface.co/linoyts/ltx2.3-inpainting-lora)
+on [`diffusers/LTX-2.3-Distilled-Diffusers`](https://huggingface.co/diffusers/LTX-2.3-Distilled-Diffusers)
+via `LTX2InContextPipeline`. 8-step distilled schedule, LoRA scale 1.0.