from PIL import Image
import torch

image = Image.open("earth.jpg")

inputs = processor(images=image, return_tensors="pt")
outputs = model(**inputs)
pred = outputs.logits.argmax(-1).item()

print(model.config.id2label[pred])