Update app.py

app.py

@@ -10,7 +10,9 @@ from langchain.chains import ConversationalRetrievalChain
 from htmlTemplates import css, bot_template, user_template
 from langchain_community.llms import HuggingFaceHub
 import os
-from sentence_transformers import SentenceTransformer
+# from sentence_transformers import SentenceTransformer
+from langchain_huggingface import HuggingFaceEmbeddings
+
 
 # from huggingface_hub import login
 
@@ -41,17 +43,22 @@ def get_text_chunks(text):
 #   vectorstore = FAISS.from_texts(texts=text_chunks, embedding=embeddings)
 #   return vectorstore
 
-def get_vectorstore(text_chunks):
-    # Load a SentenceTransformer model for embeddings
-    embedding_model = SentenceTransformer("hkunlp/instructor-xl")  # Replace with a model of your choice
-    embeddings = [embedding_model.encode(chunk) for chunk in text_chunks]
+# def get_vectorstore(text_chunks):
+#     # Load a SentenceTransformer model for embeddings
+#     embedding_model = SentenceTransformer("hkunlp/instructor-xl")  # Replace with a model of your choice
+#     embeddings = [embedding_model.encode(chunk) for chunk in text_chunks]
 
-    # Create a FAISS vectorstore
-    vectorstore = FAISS.from_embeddings(embeddings=embeddings, texts=text_chunks)
-    return vectorstore
+#     # Create a FAISS vectorstore
+#     vectorstore = FAISS.from_embeddings(embeddings=embeddings, texts=text_chunks)
+#     return vectorstore
 
+def get_vectorstore(text_chunks):
+  embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
+  vectorstore = FAISS.from_texts(texts=text_chunks, embedding=embeddings)
+  return vectorstore
+    
 def get_conversation_chain(vectorstore):
-  llm = HuggingFaceHub(repo_id="google/flan-t5-xxl", model_kwargs={"temperature":0.5, "max_length":512})
+  llm = HuggingFaceHub(repo_id="google/flan-t5-xxl", model_kwargs={"temperature":0.5, "max_length":512},huggingfacehub_api_token=os.getenv("TOKEN_API2"))
   memory=ConversationBufferMemory(
       memory_key='chat_history',return_messages=True)
   conversation_chain = ConversationalRetrievalChain.from_llm(