Added the title generation functionality

examples/openai-bot/main.py

@@ -1,25 +1,72 @@
+from langchain import LLMChain
 from textbase import bot, Message
 from textbase.models import OpenAI
 from typing import List
+from langchain.llms import OpenAI as LangOpenAI
+from langchain.prompts import PromptTemplate
+from langchain.memory import ConversationBufferMemory
+from langchain.utilities import WikipediaAPIWrapper
 
-# Load your OpenAI API key
+# Set your LangChain API key
+langchain_api_key = 'YOUR_LANGCHAIN_API_KEY'
+
+# Initialize LangChain components
+llm = LangOpenAI(api_key=langchain_api_key, temperature=0.9)
+title_template = PromptTemplate(
+    input_variables=['topic'],
+    template='Get me important notes from youtube for the topic known as {topic}'
+)
+title_memory = ConversationBufferMemory(
+    input_key='topic', memory_key='chat_history'
+)
+wiki = WikipediaAPIWrapper()
+
+# Load your OpenAI API key for TextBase
 OpenAI.api_key = ""
 
 # Prompt for GPT-3.5 Turbo
-SYSTEM_PROMPT = """You are chatting with an AI. There are no specific prefixes for responses, so you can ask or talk about anything you like.
-The AI will respond in a natural, conversational manner. Feel free to start the conversation with any question or topic, and let's have a
-pleasant chat!
+SYSTEM_PROMPT = """
+Welcome to the YouTube Content and Title Generator Chatbot!
+I'm here to help you come up with creative YouTube video ideas and titles.
+Please choose one of the following options or provide specific prompts:
+1. Generate video ideas based on a topic or keyword.
+2. Suggest catchy video titles for your content.
+3. Combine multiple topics for unique video concepts.
+4. Request video ideas tailored to your audience.
+5. Ask for tips on optimizing YouTube video titles and descriptions.
+6. Something else (Feel free to ask any YouTube-related question or request).
 """
 
 @bot()
 def on_message(message_history: List[Message], state: dict = None):
+    user_message = message_history[-1].text
 
-    # Generate GPT-3.5 Turbo response
-    bot_response = OpenAI.generate(
-        system_prompt=SYSTEM_PROMPT,
-        message_history=message_history, # Assuming history is the list of user messages
-        model="gpt-3.5-turbo",
-    )
+    # Check if the user requests YouTube content or titles
+    if "generate YouTube content" in user_message.lower():
+        # Extract the topic from the user's message
+        topic = user_message.split("generate YouTube content")[1].strip()
+
+        # Call LangChain to generate YouTube content
+        title_chain = LLMChain(llm=llm, prompt=title_template,
+                               verbose=True, output_key='title', memory=title_memory)
+        script_chain = LLMChain(llm=llm, prompt=script_template,
+                                verbose=True, output_key='script', memory=script_memory)
+
+        # Generate title and research using LangChain
+        title = title_chain.run({"topic": topic})
+        wiki_research = wiki.run({"topic": topic})
+
+        # You can format and return the generated content as needed
+        generated_content = f"Title: {title}\nWikipedia Research: {wiki_research}"
+
+        bot_response = generated_content
+    else:
+        # Generate GPT-3.5 Turbo response based on SYSTEM_PROMPT
+        bot_response = OpenAI.generate(
+            system_prompt=SYSTEM_PROMPT,
+            message_history=message_history,  # Assuming history is the list of user messages
+            model="gpt-3.5-turbo",
+        )
 
     response = {
         "data": {
@@ -41,4 +88,4 @@ def on_message(message_history: List[Message], state: dict = None):
     return {
         "status_code": 200,
         "response": response
-    }
+    }