Update app.py

app.py

@@ -24,18 +24,19 @@ QA_CHAIN_PROMPT = PromptTemplate(input_variables = ["context", "question"], temp
 
 def invoke(openai_api_key, youtube_url, prompt):
     openai.api_key = openai_api_key
-    youtube_dir = "docs/youtube/"
-    loader = GenericLoader(YoutubeAudioLoader([youtube_url], youtube_dir), OpenAIWhisperParser())
-    docs = loader.load()
-    text_splitter = RecursiveCharacterTextSplitter(chunk_size = 1500, chunk_overlap = 150)
-    splits = text_splitter.split_documents(docs)
-    chroma_dir = "docs/chroma/"
-    vectordb = Chroma.from_documents(documents = splits, embedding = OpenAIEmbeddings(), persist_directory = chroma_dir)
-    llm = ChatOpenAI(model_name = "gpt-4", temperature = 0)
-    qa_chain = RetrievalQA.from_chain_type(llm, retriever = vectordb.as_retriever(), return_source_documents = True, chain_type_kwargs = {"prompt": QA_CHAIN_PROMPT})
+    if vectordb == None:
+        youtube_dir = "docs/youtube/"
+        loader = GenericLoader(YoutubeAudioLoader([youtube_url], youtube_dir), OpenAIWhisperParser())
+        docs = loader.load()
+        text_splitter = RecursiveCharacterTextSplitter(chunk_size = 1500, chunk_overlap = 150)
+        splits = text_splitter.split_documents(docs)
+        chroma_dir = "docs/chroma/"
+        vectordb = Chroma.from_documents(documents = splits, embedding = OpenAIEmbeddings(), persist_directory = chroma_dir)
+        llm = ChatOpenAI(model_name = "gpt-4", temperature = 0)
+        qa_chain = RetrievalQA.from_chain_type(llm, retriever = vectordb.as_retriever(), return_source_documents = True, chain_type_kwargs = {"prompt": QA_CHAIN_PROMPT})
     result = qa_chain({"query": prompt})
-    shutil.rmtree(youtube_dir)
-    shutil.rmtree(chroma_dir)
+    #shutil.rmtree(youtube_dir)
+    #shutil.rmtree(chroma_dir)
     return result["result"]
 
 description = """The app demonstrates how to use a <strong>Large Language Model</strong> (LLM) with <strong>Retrieval Augmented Generation</strong> (RAG) on external data. 
@@ -47,7 +48,7 @@ description = """The app demonstrates how to use a <strong>Large Language Model<
 
 gr.close_all()
 demo = gr.Interface(fn=invoke, 
-                    inputs = [gr.Textbox(label = "OpenAI API Key", value = "sk-", lines = 1), gr.Textbox(label = "YouTube URL", value = "https://www.youtube.com/watch?v=--khbXchTeE", lines = 1), gr.Textbox(label = "Prompt", value = "What is GPT-4?", lines = 1)],
+                    inputs = [gr.Textbox(label = "OpenAI API Key", value = "sk-", lines = 1), gr.Textbox(label = "YouTube URL", value = "https://www.youtube.com/watch?v=--khbXchTeE", lines = 1), gr.Textbox(label = "Prompt", value = "GPT-4 human level performance", lines = 1)],
                     outputs = [gr.Textbox(label = "Completion", lines = 1)],
                     title = "Generative AI - LLM & RAG",
                     description = description)