Merge pull request #15 from almutareb/save_conversations

app.py

@@ -6,10 +6,38 @@ from innovation_pathfinder_ai.source_container.container import (
 from innovation_pathfinder_ai.utils.utils import extract_urls
 from innovation_pathfinder_ai.utils import logger
 
+from innovation_pathfinder_ai.utils.utils import (
+    generate_uuid    
+)
+from langchain_community.vectorstores import Chroma
+
+import chromadb
+import dotenv
+import os
+
+dotenv.load_dotenv()
+
 logger = logger.get_console_logger("app")
 
+def initialize_chroma_db() -> Chroma:
+    collection_name=os.getenv("CONVERSATION_COLLECTION_NAME")
+    
+    client = chromadb.PersistentClient()
+    
+    collection = client.get_or_create_collection(
+    name=collection_name,
+    )
+    
+    return collection
+
+
+
 if __name__ == "__main__":
     
+    current_id = generate_uuid()
+    
+    db = initialize_chroma_db()
+    
     def add_text(history, text):
         history = history + [(text, None)]
         return history, ""
@@ -30,6 +58,18 @@ if __name__ == "__main__":
                 "chat_history": history
             }
         )
+        
+        db.add(
+            ids=[current_id],
+            documents=[result['output']],
+            metadatas=[
+                {
+                    "query":query,
+                    "intermediate_steps":result['intermediate_steps'].__str__()
+                }
+            ]
+        )
+        
         return result
 
     def vote(data: gr.LikeData):

example.env

@@ -8,4 +8,6 @@ OLLMA_BASE_URL=
 SERPAPI_API_KEY=
 
 # for chromadb
-VECTOR_DATABASE_LOCATION=
+VECTOR_DATABASE_LOCATION=
+
+CONVERSATION_COLLECTION_NAME="ConversationMemory"

innovation_pathfinder_ai/structured_tools/structured_tools.py

@@ -94,6 +94,7 @@ def wikipedia_search(query: str) -> str:
 @tool
 def chroma_search(query:str) -> str:
     """Search the Arxiv vector store for docmunets and relevent chunks"""
+    # Since we have more than one collections we should change the name of this tool
     client = chromadb.PersistentClient(
     # path=persist_directory,
     )
@@ -155,4 +156,29 @@ def embed_arvix_paper(paper_id:str) -> None:
         collection_name=collection_name,
         pdf_file_location=full_path,
     )
-    
+    
+@tool
+def conversational_search(query:str) -> str:
+    """Search from past conversations  for docmunets and relevent chunks"""
+    # Since we have more than one collections we should change the name of this tool
+    client = chromadb.PersistentClient(
+    # path=persist_directory,
+    )
+    
+    collection_name=os.getenv("CONVERSATION_COLLECTION_NAME")
+    #store using envar
+    
+    embedding_function = SentenceTransformerEmbeddings(
+        model_name="all-MiniLM-L6-v2",
+        )
+    
+    vector_db = Chroma(
+    client=client, # client for Chroma
+    collection_name=collection_name,
+    embedding_function=embedding_function,
+    )
+    
+    retriever = vector_db.as_retriever()
+    docs = retriever.get_relevant_documents(query)
+    
+    return docs.__str__()

innovation_pathfinder_ai/utils/utils.py

@@ -1,6 +1,7 @@
 import hashlib
 import datetime
 import os
+import uuid
 
 from innovation_pathfinder_ai.utils import logger
 
@@ -182,4 +183,13 @@ def create_folder_if_not_exists(folder_path: str) -> None:
         os.makedirs(folder_path)
         print(f"Folder '{folder_path}' created.")
     else:
-        print(f"Folder '{folder_path}' already exists.")
+        print(f"Folder '{folder_path}' already exists.")
+        
+def generate_uuid() -> str:
+    """
+    Generate a UUID (Universally Unique Identifier) and return it as a string.
+
+    Returns:
+        str: A UUID string.
+    """
+    return str(uuid.uuid4())        

innovation_pathfinder_ai/vector_store/chroma_vector_store.py

@@ -21,8 +21,10 @@ from langchain_community.vectorstores import Chroma
 from langchain_community.embeddings.sentence_transformer import (
     SentenceTransformerEmbeddings,
 )
+from innovation_pathfinder_ai.utils.utils import (
+    generate_uuid    
+)
 
-import uuid
 import dotenv
 import os
 
@@ -31,14 +33,7 @@ dotenv.load_dotenv()
 
 VECTOR_DATABASE_LOCATION = os.getenv("VECTOR_DATABASE_LOCATION")
 
-def generate_uuid() -> str:
-    """
-    Generate a UUID (Universally Unique Identifier) and return it as a string.
 
-    Returns:
-        str: A UUID string.
-    """
-    return str(uuid.uuid4())
 
 def read_markdown_file(file_path: str) -> str:
     """