Update app.py

app.py

@@ -9,18 +9,27 @@ import gradio as gr
 import pickle
 from threading import Lock
 
-model_sbert = "sentence-transformers/all-mpnet-base-v2"
+model_options = {'all-mpnet-base-v2': "sentence-transformers/all-mpnet-base-v2",
+                'instructor-base'}: "hkunlp/instructor-base"}
 
+def load_vectorstore(model):
+    '''load embeddings and vectorstore'''
 
-sbert_emb = HuggingFaceEmbeddings(model_name=model_sbert)
+    if 'mpnet' in model:
+        
+        emb = HuggingFaceEmbeddings(model_name=model)
 
-def load_vectorstore(folder,embeddings):
-    
-    vectorstore = FAISS.load_local(folder,embeddings) 
+        return FAISS.load_local('vanguard-embeddings', emb)
+
+    elif 'instructor'in model:
+        
+        emb = HuggingFaceInstructEmbeddings(model_name=model,
+                                               query_instruction='Represent the Financial question for retrieving supporting paragraphs: ',
+                                               embed_instruction='Represent the Financial paragraph for retrieval: ')
+        return FAISS.load_local('vanguard-embeddings-inst', emb)
 
-    return vectorstore
 
-vectorstore = load_vectorstore('vanguard-embeddings',sbert_emb)
+# vectorstore = load_vectorstore('vanguard-embeddings',sbert_emb)
     
 _template = """Given the following conversation and a follow up question, rephrase the follow up question to be a standalone question.
 You can assume the question about investing and the investment management industry.
@@ -104,6 +113,10 @@ with block:
             type="password",
         )
 
+    embeddings = gr.Radio(choices=list(model_options.keys()),value=list(model_options.keys())[0], label='Choose your Embedding Model')
+
+    vectorstore = load_vectorstore(model_options[embeddings])
+
     chatbot = gr.Chatbot()
 
     with gr.Row():
@@ -133,6 +146,7 @@ with block:
     state = gr.State()
     agent_state = gr.State()
 
+    
     submit.click(chat, inputs=[openai_api_key_textbox, message, state, agent_state], outputs=[chatbot, state])
     message.submit(chat, inputs=[openai_api_key_textbox, message, state, agent_state], outputs=[chatbot, state])