Merge pull request #1 from Pushkar222-n/feature/GraphDB-search

config.py

@@ -0,0 +1,18 @@
+from pydantic_settings import BaseSettings
+import os
+
+
+class Settings(BaseSettings):
+    groq_api_key: str
+    neo4j_uri: str
+    neo4j_username: str
+    neo4j_password: str
+    neo4j_database: str
+    aura_instanceid: str
+    aura_instancename: str
+
+    class Config:
+        env_file = '.env'
+
+
+settings = Settings()  # type: ignore

src/api/main.py

@@ -78,7 +78,7 @@ async def get_recommendations(request: RecommendationRequest):
             recommendations=result["recommendations"],
             retrieved_count=result["retrieved_count"],
             metadata={
-                "model": "llama-3.3-70b-versatile",
+                "model": "llama-3.1-8b-instant",
                 "retriever_k": rag_pipeline.retriever_k,
                 "Time taken for LLM + vector search": str(end_time - start_time)
             }
@@ -97,7 +97,7 @@ async def get_stats():
     return {
         "total_anime": rag_pipeline.retriever.collection.count(),
         "embedding_model": "all-MiniLM-L6-v2",
-        "llm_model": "llama-3.1-70b-versatile",
+        "llm_model": "llama-3.1-8b-instant",
         "retrieval_k": rag_pipeline.retriever_k
     }
 

src/data_ingestion/load_to_neo4j.py

@@ -0,0 +1,24 @@
+from neo4j import GraphDatabase
+import pandas as pd
+from src.utils.neo4j_client import Neo4j_Client
+
+class AnimeGraphLoader:
+    """Loads anime into Neo4j Graph"""
+
+    def __init__(self):
+        self.client = Neo4j_Client()
+
+    def clear_database(self):
+        """Delete all nodes (use carefully !)"""
+        query = "MATCH (n) DETACH DELETE n"
+        self.client.run_query(query)
+    
+    def create_constraints(self):
+        """Create uniqueness constraints for performance"""
+        queries = [
+            "CREATE CONSTRAINT anime_id IF NOT EXISTS FOR (a: Anime) REQUIRE a.mal_id IS UNIQUE",
+            "CREATE CONSTRAINT genre_name IF NOT EXISTS FOR (g:Genre) REQUIRE g.name IS UNIQUE",
+            "CREATE CONSTRAINT theme_name IF NOT EXISTS FOR (t:Theme) REQUIRE t.name IS UNIQUE",
+            "CREATE CONSTRAINT demographic_name IF NOT EXISTS FOR (d:Demographic) REQUIRE d.name IS UNIQUE",
+            "CREAE CONSTRAINT content_type IF NOT EXISTS FOR (t: Type) REQUIRE t.type IS UNIUE"
+        ]

src/llm/groq_client.py

@@ -2,8 +2,8 @@ import os
 from groq import Groq
 from dotenv import load_dotenv
 import logging
+from config import settings
 
-load_dotenv()
 logger = logging.getLogger(__name__)
 
 
@@ -19,7 +19,7 @@ class GroqLLM:
         - llama-3.1-8b-instant: Faster, good enough for most tasks
         """
 
-        self.client = Groq()
+        self.client = Groq(api_key=settings.groq_api_key)
         self.model = model
 
         logger.info(f"Initialized Groq with model {self.model}")

src/retrieval/rag_pipeline.py

@@ -24,7 +24,7 @@ class AnimeRAGPipeline:
             recommendation_n: How many to recommend in final output
         """
         self.retriever = retriever or AnimeRetriever()
-        self.llm = llm or GroqLLM(model="llama-3.3-70b-versatile")
+        self.llm = llm or GroqLLM(model="llama-3.1-8b-instant")
         self.retriever_k = retriever_k
         self.recommendation_n = recommendation_n
 

src/utils/neo4j_client.py

@@ -1,17 +1,16 @@
-import os
 from dotenv import load_dotenv
 from neo4j import GraphDatabase
 
-load_dotenv()
+from config import settings
 
 
 class Neo4j_Client:
     """Neo4j database client"""
 
     def __init__(self):
-        uri = os.getenv("NEO4J_URI")
-        user = os.getenv("NEO4J_USERNAME")
-        password = os.getenv("NEO4J_PASSWORD")
+        uri = settings.neo4j_uri
+        user = settings.neo4j_username
+        password = settings.neo4j_password
         try:
             self.driver = GraphDatabase.driver(
                 uri, auth=(user, password))  # type: ignore