diff --git a/emotion-based music recommendtion/back.py b/emotion-based music recommendtion/back.py
index 11d6b76bba..7738e0b6e1 100644
--- a/emotion-based music recommendtion/back.py	
+++ b/emotion-based music recommendtion/back.py	
@@ -4,6 +4,8 @@
 # YouTube API credentials
 api_key = 'Your- API- Key'
 
+
+# Initialize emotion classifier
 model_name = 'bhadresh-savani/bert-base-go-emotion'
 model = AutoModelForSequenceClassification.from_pretrained(model_name, from_tf=False)
 tokenizer = AutoTokenizer.from_pretrained(model_name)
@@ -29,7 +31,7 @@ def map_emotion_to_query(emotion):
     else:
         return 'relaxing music'
 
-def get_youtube_recommendations(query, language=None, order='relevance'):
+def get_youtube_recommendations(query, language=None):
     youtube = googleapiclient.discovery.build('youtube', 'v3', developerKey=api_key)
     if language and language.lower() != 'i am feeling lucky':
         query = f"{query} {language} songs"
@@ -39,8 +41,7 @@ def get_youtube_recommendations(query, language=None, order='relevance'):
         q=query,
         type='video',
         videoDuration='medium',
-        videoCategoryId='10',
-        order=order
+        videoCategoryId='10'
     )
     search_response = search_request.execute()
     video_ids = [item['id']['videoId'] for item in search_response['items']]
@@ -51,4 +52,4 @@ def get_youtube_recommendations(query, language=None, order='relevance'):
     video_response = video_request.execute()
     videos = sorted(video_response['items'], key=lambda x: int(x['statistics'].get('likeCount', 0)), reverse=True)
     top_videos = [f"{item['snippet']['title']} (https://www.youtube.com/watch?v={item['id']}) with {item['statistics'].get('likeCount', 0)} likes" for item in videos[:5]]
-    return top_videos if top_videos else ["No recommendations available for this emotion and language"]
\ No newline at end of file
+    return top_videos if top_videos else ["No recommendations available for this emotion and language"]