src.rag: introduced FAISS option for retriever, and made it default

src/rag.py

@@ -6,6 +6,7 @@ from langchain.text_splitter import RecursiveCharacterTextSplitter
 
 from langchain_huggingface import HuggingFaceEmbeddings
 from langchain_community.vectorstores import Chroma
+from langchain_community.vectorstores import FAISS
 from langchain_openai import ChatOpenAI
 # from langchain_community.llms import HuggingFaceHub
 from langchain_huggingface import HuggingFaceEndpoint
@@ -28,6 +29,10 @@ class RAG():
         # self.use_model = 'gpt-4o-mini'
         self.use_model = 'zephyr-7b-alpha'
 
+        # self.use_vectordb = 'chroma'
+        self.use_vectordb = 'faiss'
+
+
         # Load environment variables that should contain:
         #  - 'OPENAI_API_KEY' for OpenAI models
         #  - 'HUGGINGFACEHUB_API_TOKEN' for HuggingFace models
@@ -71,9 +76,14 @@ class RAG():
 
     def create_retriever(self, texts, embeddings):
         # Create embeddings and vector store
-        print ('Creating vectore store with Chroma')
-        vectorstore = Chroma.from_documents(texts, embeddings)
-        retriever = vectorstore.as_retriever(search_kwargs={"k": self.k})
+        if self.use_vectordb == 'chroma':
+            print ('Creating vectore store with Chroma')
+            vectorstore = Chroma.from_documents(texts, embeddings)
+            retriever = vectorstore.as_retriever(search_kwargs={"k": self.k})
+        elif self.use_vectordb == 'faiss':
+            print ('Creating vectore store with FAISS')
+            vectorstore = FAISS.from_documents(texts, embeddings)
+            retriever = vectorstore.as_retriever(search_kwargs={"k": self.k})
         return retriever
 
     def create_llm(self):
@@ -84,7 +94,7 @@ class RAG():
                 model_name="gpt-4o-mini",
                 temperature=0)
         elif self.use_model ==  'zephyr-7b-alpha':
-            print(f'As llm, using HF model: {self.use_model}')
+            print(f'As llm, using HF-Endpint: {self.use_model}')
             llm = HuggingFaceEndpoint(
                 repo_id=f"huggingfaceh4/{self.use_model}",
                 temperature=0.1,