Spaces:

naver-ai
/

VisualStylePrompting_Controlnet

Running on Zero

taki0112 commited on Mar 25

Commit

aab4477

•

1 Parent(s): 56f7912

fix

Files changed (1) hide show

app.py CHANGED Viewed

@@ -226,7 +226,6 @@ description_md = """
 ### 🔥 [[Default ver](https://huggingface.co/spaces/naver-ai/VisualStylePrompting)]
 ---
 ### ✨ Visual Style Prompting also works on `ControlNet` which specifies the shape of the results by depthmap or keypoints.
-### ‼️ w/ ControlNet ver does not support user style images.
 ### 🔥 To try out our demo with ControlNet,
 1. Upload an `image for depth control`. An off-the-shelf model will produce the depthmap from it.
 2. Choose `ControlNet scale` which determines the alignment to the depthmap.
@@ -234,6 +233,7 @@ description_md = """
 4. Enter the `text prompt`. (`Empty text` is okay, but a depthmap description helps.)
 5. Choose the `number of outputs`.
 ### 👉 To achieve faster results, we recommend lowering the diffusion steps to 30.
 ### Enjoy ! 😄
 """

 ### 🔥 [[Default ver](https://huggingface.co/spaces/naver-ai/VisualStylePrompting)]
 ---
 ### ✨ Visual Style Prompting also works on `ControlNet` which specifies the shape of the results by depthmap or keypoints.
 ### 🔥 To try out our demo with ControlNet,
 1. Upload an `image for depth control`. An off-the-shelf model will produce the depthmap from it.
 2. Choose `ControlNet scale` which determines the alignment to the depthmap.
 4. Enter the `text prompt`. (`Empty text` is okay, but a depthmap description helps.)
 5. Choose the `number of outputs`.
+### ⚠️ w/ ControlNet ver does not support user style images.
 ### 👉 To achieve faster results, we recommend lowering the diffusion steps to 30.
 ### Enjoy ! 😄
 """