adding Prompt

app.py

@@ -5,7 +5,6 @@ from langchain.document_loaders import PDFMinerLoader,CSVLoader ,UnstructuredWor
 from langchain.text_splitter import CharacterTextSplitter
 from langchain.embeddings import SentenceTransformerEmbeddings
 from langchain.vectorstores import FAISS
-from langchain import HuggingFaceHub
 from langchain.chains import RetrievalQA
 from langchain.prompts import PromptTemplate
 from langchain.docstore.document import Document
@@ -51,6 +50,20 @@ def process_youtube_link(link, document_name="youtube-content"):
         logger.error(f'Error in reading document. {err}')
 
 
+def create_prompt():
+    prompt_template = """As a chatbot asnwer the questions regarding the content in the video. 
+    Use the following context to answer. 
+    If you don't know the answer, just say I don't know. 
+
+    {context}
+
+    Question: {question}
+    Answer :"""
+    prompt = PromptTemplate(
+        template=prompt_template, input_variables=["context", "question"]
+    )
+    return prompt
+
 def youtube_chat(youtube_link,API_key,llm='HuggingFace',temperature=0.1,max_tokens=1096,char_length=1500):
     
     document  = process_youtube_link(link=youtube_link)
@@ -69,11 +82,12 @@ def youtube_chat(youtube_link,API_key,llm='HuggingFace',temperature=0.1,max_toke
                             )
     else:
         chat = chatops.get_openai_chat_model(API_key=API_key)
+    chain_type_kwargs = {"prompt": create_prompt()}
 
     qa = RetrievalQA.from_chain_type(llm=chat,
                                 chain_type='stuff',
                                 retriever=vector_db.as_retriever(),
-                                #  chain_type_kwargs=chain_type_kwargs,
+                                chain_type_kwargs=chain_type_kwargs,
                                 return_source_documents=True
                             )
     return "Youtube link Processing completed ..."
@@ -85,7 +99,6 @@ def infer(question, history):
     # #     res.append(pair)
     
     # chat_history = res
-    print("Question in infer :",question)
     result = qa({"query": question})
     matching_docs_score = vector_db.similarity_search_with_score(question)