Update app.py

app.py

@@ -4,6 +4,7 @@ from transformers import  ViTImageProcessor, VisionEncoderDecoderModel,RobertaTo
 import PIL 
 import streamlit as st
 from PIL import Image
+import trasformer
 
 
 def set_page_config():
@@ -16,9 +17,9 @@ def set_page_config():
 def initialize_model():
     device = 'cpu'
     config = AutoConfig.from_pretrained("sourabhbargi11/Caption_generator_model")  
-    model = VisionEncoderDecoderModel.from_pretrained("sourabhbargi11/Caption_generator_model", config=config).to(device)
+    model = VisionEncoderDecoderModel.from_pretrained("sourabhbargi11/Caption_generator_model", config=config)
     tokenizer = RobertaTokenizerFast.from_pretrained("roberta-base")
-    image_processor = ViTImageProcessor.from_pretrained("google/vit-base-patch16-224",device=device)
+    image_processor = ViTImageProcessor.from_pretrained("google/vit-base-patch16-224")
     return image_processor, model,tokenizer, device
 
 def upload_image():
@@ -31,7 +32,7 @@ def image_preprocess(image):
     return image
 
 def generate_caption(image_processor, model, tokenizer,device, image):
-    inputs = image_processor(image, return_tensors='pt').to(device)
+    inputs = image_processor(image, return_tensors='pt')
     print(inputs)
     model.eval()
     # Generate caption