Hello everyone! I’ve converted my CLIP model to ONNX, and then wanted to run an inference, but it doesn’t work without text in input… Is it possible to fix that?