Removed unnecessary lines

app/vectorizer.py

@@ -9,10 +9,6 @@ from pinecone import Pinecone, ServerlessSpec
 # Disable parallelism for tokenizers
 os.environ['TOKENIZERS_PARALLELISM'] = 'false'
 
-# Ensure the cache directory exists
-os.makedirs('/app/cache/hub', exist_ok=True)
-os.environ['HF_HOME'] = '/app/cache/hub'
-
 # Configure logging
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
@@ -51,6 +47,7 @@ class Vectorizer:
                 )
         else:
             logger.info(f"Pinecone index {self.pinecone_index_name} already exists")
+            pinecone.delete_index(self.pinecone_index_name)
 
         self.index = pinecone.Index(self.pinecone_index_name)
 

load_data.py

@@ -1,8 +1,6 @@
 import os
 from app.vectorizer import Vectorizer
 
-os.environ['HF_HOME'] = '/app/cache/hub'
-
 if __name__ == "__main__":
     vectorizer = Vectorizer()
     vectorizer.store_from_dataset(store_data=True)  # Run this once to load the dataset into Pinecone

ui/app.py

@@ -14,8 +14,8 @@ n = st.number_input("Number of results:", min_value=1, max_value=20, value=5)
 
 if st.button("Search"):
     search_method = use_pinecone == 'Pinecone Vector Search'
-    backend_url = "http://supertskone-prompt-search-engine.hf.space/search" 
-    response = requests.post(backend_url, json={"query": query, "n": n, "use_pinecone": search_method})
+    response = requests.post("http://localhost:5000/search",
+                             json={"query": query, "n": n, "use_pinecone": search_method})
 
     # Log the response for debugging
     st.write("Response Status Code:", response.status_code)