What happens if I exceed the number of tokens that fit in the widget?
I tested the model using CLIP on the web widget.
I put the long prompt to generate the image, and even the last part of the prompt seemed to be reflected in the result.
No error messages were thrown, and there was no indication of how excess tokens would be handled.
However, when I try to test the model on local environment(using diffusers or safetensors), in the same prompt it warns me that I have exceeded 77 tokens.
If anyone knows exactly how this works in the web widget, please let me know.