After startup, stay on the Loading... page

Question

After startup, stay on the Loading... page

supeeeeee opened this issue a year ago · 5 comments

Answer 1 · 2023-05-08T04:45:24.000Z

Have you put the weights files (SAM models and OWL-ViT models) into the checkpoints folder ? Or do you have any other error information?

Answer 2 · 2023-05-08T07:48:39.000Z

I think all 3 SAM models and the OWL-VIT model("pytorch_model.bin" ?) are in the right place, the app.py file can run, but after startup and click the local URL the page stay on the Loading....
The only UserWarning is : Expected at least 14 arguments for function <function run_inference at 0x7f06a9fa1820>, received 13.

Answer 3 · 2023-05-08T10:15:38.000Z

OWL-ViT model should include the above files. And please check the Python packages version in requirements.txt.

Answer 4 · 2023-05-10T03:09:07.000Z

Thanks a lot bro!! but I still have a problem:

"/anaconda3/envs/SAM/lib/python3.10/site-packages/gradio_client/serializing.py", line 41, in input_api_info
return (api_info["serialized_input"][0], api_info["serialized_input"][1])
KeyError: 'serialized_input'"

I think my Gradio version is right(gradio==3.25.0), but I still get this warning.

Answer 5 · 2023-05-10T14:13:51.000Z

Thanks a lot bro!! but I still have a problem:

"/anaconda3/envs/SAM/lib/python3.10/site-packages/gradio_client/serializing.py", line 41, in input_api_info return (api_info["serialized_input"][0], api_info["serialized_input"][1]) KeyError: 'serialized_input'"

I think my Gradio version is right(gradio==3.25.0), but I still get this warning.

I think the function run_inference in inference.py has a bug. The param selected_points should be selected_points=[]. I have pushed the fixed code.