gemini added

.env

@@ -0,0 +1 @@
+Google_api = 'AIzaSyB3wI2r6ZgQnYQ3V39PX5S0zWSRqy5ldYw'

main.py

@@ -4,7 +4,7 @@ from utils.convert_embedding import GetEmbedding
 import random
 import pickle
 import os 
-
+from utils.rag import RAG
 
 
 # def dump_user_question(query):
@@ -35,10 +35,10 @@ def process(user_query:str):
     score = similarity_scores[0,index]
     print(f"Index : {index}:\tscore:{score} \tquery: {user_query}")
 
-    if float(score) > 0.45 :
+    if float(score) > 0.60 :
         final_output = random.choice(answer)
     else:
-        final_output = "Sorry, I did not understand."
+        final_output = RAG().pipeline(query=user_query)
 
     return final_output
 

utils/rag.py

@@ -0,0 +1,42 @@
+from langchain_google_genai import GoogleGenerativeAI
+import requests
+from bs4 import BeautifulSoup
+from dotenv import load_dotenv
+import os
+load_dotenv()
+
+
+class RAG:
+    def __init__(self):
+        self.url = 'https://lalitmahale.github.io'
+        self.llm = GoogleGenerativeAI(google_api_key=os.getenv("Google_api"),model="gemini-pro")
+
+    
+    def get_data(self):
+        try:
+            res = requests.get(self.url)
+            soup = BeautifulSoup(res.content, "html.parser")
+            return soup.get_text()
+        except Exception as e:
+            print(e)
+
+    def clean_text(self):
+        return self.get_data().replace("\n","")
+
+    def prompt(self):
+        return """You are a helpfull assistant for me. understand the below context and give answer for user question.
+        
+        context : {context}\n\nQuestion : {question}\n\nGive proper answer for this questions."""
+
+
+    def pipeline(self,query):
+        try:
+            prompt = self.prompt().format(context = self.clean_text(),question = query)
+            return self.llm.invoke(prompt)
+        except Exception as e:
+            print(e)
+    
+
+if __name__ == "__main__":
+    res = RAG().pipeline("who is lalit mahale")
+    print(res)