Update app.py

app.py

@@ -29,6 +29,10 @@ inputs = processor(
     return_tensors="pt",
 ).to(device)
 
+# Check the number of tokens generated by the processor and the shape of inputs
+print("Input tokens:", inputs.input_ids.shape)
+print("Image features shape:", inputs.pixel_values.shape)
+
 # Inference
 generated_ids = model.generate(**inputs, max_new_tokens=128)