# pass image and text classes to processor
inputs = processor(text=text_classes, images=image, return_tensors="pt", 
                                                    padding=True)

# pass inputs to CLIP
outputs = model(**inputs) # note: "**" unpacks dictionary items