Update app.py

app.py

@@ -31,7 +31,7 @@ def image_preprocess(image):
     return image
 
 def generate_caption(image_processor, model, tokenizer,device, image):
-    inputs = image_processor(image,batch_size=1, return_tensors='pt').to(device)
+    inputs = image_processor(image, return_tensors='pt').to(device)
     model.eval()
     # Generate caption
     with torch.no_grad():
@@ -40,6 +40,7 @@ def generate_caption(image_processor, model, tokenizer,device, image):
           max_length=1000,# Adjust the maximum length of the generated caption as needed
           num_beams=4,    # Adjust the number of beams for beam search decoding
           early_stopping=True  # Enable early stopping to stop generation when all beams finished
+          ,batch_size=1
       )
 
       # Decode the generated caption