Update app.py

app.py

@@ -599,16 +599,17 @@ def load_and_preprocess_image(img):
     return np.expand_dims(img, axis=0)
 
 
-def generate_caption_coca(image):
+def generate_caption(image):
     img_processed = load_and_preprocess_image(image)
     img_processed = np.expand_dims(img_processed, axis=0)
 
     _, cap_features = coca_model.encoder.predict(img_processed, verbose=0)
+    cap_features = cap_features.astype(np.float32) 
 
     start_token_id = word_index[start_token]
     end_token_id = word_index[end_token]
     sequence = [start_token_id]
-    text_input = np.zeros((1, sentence_length - 1))
+    text_input = np.zeros((1, sentence_length - 1), dtype=np.float32)
 
     for t in range(sentence_length - 1):
         text_input[0, :len(sequence)] = sequence
@@ -630,6 +631,7 @@ def generate_caption_coca(image):
 
     return caption
     
+    
 # def generate_caption_coca(image):
 #     img_processed = load_and_preprocess_image(image)
 #     _, cap_features = coca_model.encoder.predict(img_processed, verbose=0)