Not genrating response on HuggingFace Spaces
I am trying to use this space https://huggingface.co/spaces/ysharma/LLaVA_v1 Space on Huggingface? When I try to run the code it creates a JSON file, But from where to get that model output? I mean response of LLaVA? Is there any way to use this model with API?
Code:
from gradio_client import Client
client = Client("https://ysharma-llava-v1.hf.space/--replicas/nlp74/")
result = client.predict(
"What is this?", # str in 'parameter_0' Textbox component
"https://raw.githubusercontent.com/gradio-app/gradio/main/test/test_files/bus.png", # str (filepath on your computer (or URL) of image) in 'parameter_8' Image component
"Crop", # str in 'Preprocess for non-square image' Radio component
fn_index=7
)
print(result)
I encountered the same issue, and it seems that sometimes it returns an empty string. However, after a few attempts, it starts working again. I'm not sure why this happens.