Update custom_llm.py

custom_llm.py

@@ -19,6 +19,14 @@ from langchain_community.document_loaders import PyMuPDFLoader
 import os
 from langchain.embeddings import HuggingFaceEmbeddings
 from langchain.vectorstores import FAISS
+import datasets
+
+
+def create_vectorstore():
+    data = load_datasets("ruslanmv/ai-medical-chatbot", split='train')
+    emb_model = HuggingFaceEmbeddings(model_name='sentence-transformers/paraphrase-multilingual-mpnet-base-v2', encode_kwargs={'normalize_embeddings': True})
+    faiss = FAISS.from_texts(data['Doctor'], emb_model)
+    return faiss
 
 
 
@@ -26,8 +34,12 @@ def custom_chain_with_history(llm, memory):
 
     prompt = PromptTemplate.from_template("""You are a helpful, respectful, and honest assistant. Always answer as helpfully as possible while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature.
 
-    If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
+    You have the access to the following context information:
 
+    {context}
+
+    
+    If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information.
     
     {chat_history}
     ### User: {question}
@@ -44,7 +56,11 @@ def custom_chain_with_history(llm, memory):
       print(len(docs))
       return "\n".join([f"{i+1}. {d.page_content}" for i,d in enumerate(docs)])
     
-    return {"chat_history":lambda x:prompt_memory(x['memory']), "question":lambda x:x['question']} | prompt | llm
+    return {
+        "chat_history":lambda x:prompt_memory(x['memory']), 
+        "question":lambda x:x['question'], 
+        "context": itemgetter("question") | create_vectorstore().as_retriever(search_type="similarity", search_kwargs={"k": 6}) | format_docs
+    } | prompt | llm
 
 
 class CustomLLM(LLM):