which version of gradio you are using?

[moosl]

Your repo is fantastic, and I'm very grateful for your sharing. I wanted to ask, which version of Gradio are you using? I encountered some errors while running it.

[tzktz]

Your repo is fantastic, and I'm very grateful for your sharing. I wanted to ask, which version of Gradio are you using? I encountered some errors while running it.

Use gradio version 3.5..It's works fine 👍 ..Still u face any issues let me know it glad to ans it :)

[moosl]

Your repo is fantastic, and I'm very grateful for your sharing. I wanted to ask, which version of Gradio are you using? I encountered some errors while running it.

Use gradio version 3.5..It's works fine 👍 ..Still u face any issues let me know it glad to ans it :)

I am using 3.50 also, when I run the code, it comes with this error:

Traceback (most recent call last):
  File "/data/xjl/face-swap/app.py", line 76, in <module>
    gr.inputs.Image(type="file", label="Target Image"),
  File "/home/nvidia/anaconda3/envs/face-swap/lib/python3.10/site-packages/gradio/blocks.py", line 115, in wrapper
    return fn(self, **kwargs)
  File "/home/nvidia/anaconda3/envs/face-swap/lib/python3.10/site-packages/gradio/inputs.py", line 252, in __init__
    super().__init__(
  File "/home/nvidia/anaconda3/envs/face-swap/lib/python3.10/site-packages/gradio/components/image.py", line 124, in __init__
    raise ValueError(
ValueError: Invalid value for parameter `type`: file. Please choose from one of: ['numpy', 'pil', 'filepath']

Then I update the code as follows:

iface = gr.Interface(
    fn=predictor.predict,
    inputs=[
        gr.inputs.Image(type="filepath", label="Target Image"),
        gr.inputs.Image(type="filepath", label="Swap Image")
    ],
    outputs=gr.outputs.Image(type="filepath", label="Result"),
    title=title,
    examples=[["input.jpg", "swap img.jpg"]])

The program was successfully launched, but when I run it on the gradio, it had this error, there is not output on the page.

'str' object has no attribute 'name'

[moosl]

I updated the predict function, and it works now:

frame = cv2.imread(input_image) source_face = self.get_face(cv2.imread(swap_image))