Online demo's python API is raising KeyError
#1
by
gilkzxc
- opened
By using the Gradio-client API to access the ONE-PEACE model, with the Client().perdict() function.
Instead of retrieving the wanted result, for all fn_index values, a KeyError always being raised.
And this is while the Gradio-client API does work for other models on Gradio.
If the problem, lays only on the Gradio-client, is it possible to generate locally a picture from a sound/text input with ONE-PEACE API from github?
Thanks in advanced.
You can follow Multi-modal Embedding API to generate embeddings for each image in your image library. Then, recall the most relevant images based on the similarity between input text (audio) and image embeddings.