Update function.py

function.py

@@ -1,27 +1,123 @@
-from langchain_community.llms import OpenAI
-from langchain_google_genai import ChatGoogleGenerativeAI
 import streamlit as st
+from langchain.chains import LLMChain, SimpleSequentialChain
+from langchain.memory import ConversationBufferMemory
+from langchain.prompts import PromptTemplate
+from langchain.llms import OpenAI
 
 
 
-def get_answers(questions,model):
 
-
-    answer_prompt = (f"I want you to become a teacher answer this specific Question: {questions}. You should gave me a straightforward and consise explanation and answer to each one of them")
+def ai_generated_content(topic, audience, benefit, date_time):
 
     
-    if model == "Open AI":
-        llm = OpenAI(temperature=0.7, openai_api_key=st.secrets["OPENAI_API_KEY"])
-        answers = llm(answer_prompt)
-        # return questions
-        
-    elif model == "Gemini":
-        llm = ChatGoogleGenerativeAI(model="gemini-pro", google_api_key=st.secrets["GOOGLE_API_KEY"])
-        answers = llm.invoke(answer_prompt)
-        answers = answers.content
-        # return questions.content
+    
+    # Initialize the LLM and memory
+    llm = OpenAI(model="gpt-4")
+    memory = ConversationBufferMemory()
+    
+    # Define the user inputs
+    topic = topic
+    audience = audience
+    benefit_outcome = benefit
+    date = date_time
+    
+    # Store the user inputs in memory
+    memory.save_context(
+        {"input": f"Topic: {topic}, Audience: {audience}, Benefit/Outcome: {benefit_outcome}, Date: {date}"},
+        {}
+    )
+    
+   # Define prompt templates
+    headline_template = PromptTemplate(
+        input_variables=["history"],
+        template="Using the details provided: {history}, generate a headline using the format: 'MASTER [specific skill or process] IN [timeframe or tool] USING [method or tool] TO [immediate benefit] AND FINALLY [major goal or habit or benefit], GUARANTEED!'"
+    )
+    
+    subheadline_template = PromptTemplate(
+        input_variables=["history"],
+        template="Generate a subheading that reassures the user they don’t need any prior knowledge in {history}. The format should be 'WITHOUT ANY PRIOR [technical skill] OR [domain knowledge].'"
+    )
+    
+    price_statement_template = PromptTemplate(
+        input_variables=["history"],
+        template="Generate a short value statement that combines the outcome with the price of the offer. The format should be 'Become an advanced [skill] expert in [price].' Use language that emphasizes the low cost for high-value results."
+    )
+    
+    # Create individual chains
+    headline_chain = LLMChain(
+        llm=llm,
+        prompt=headline_template,
+        memory=memory,
+        output_key="headline"
+    )
+    
+    subheadline_chain = LLMChain(
+        llm=llm,
+        prompt=subheadline_template,
+        memory=memory,
+        output_key="subheadline"
+    )
+    
+    price_chain = LLMChain(
+        llm=llm,
+        prompt=price_statement_template,
+        memory=memory,
+        output_key="price_statement"
+    )
+    
+    # Combine chains into a sequential process
+    landing_page_chain = SequentialChain(
+        chains=[headline_chain, subheadline_chain, price_chain],
+        input_variables=[],
+        output_variables=["headline", "subheadline", "price_statement"]
+    )
+    
+    # Generate the landing page content
+    output = landing_page_chain.run({})
+    
+    # Parsing the outputs into key-value pairs
+    landing_page_data = {
+        "headline": output["headline"],
+        "subheadline": output["subheadline"],
+        "price_statement": output["price_statement"]
+    }
+    
+    print(landing_page_data)
+    
+    
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
 
-    return(answers)    
 
 
 
@@ -54,6 +150,17 @@ def generate_workshop_details(topic, audience, benefit, date_time):
     if not (topic and audience and benefit and date_time):
         return "All details are required to generate the workshop information."
 
+    
+    
+    
+    
+    
+    ai_generated_content(topic, audience, benefit, date_time)
+    
+    
+    
+    
+    
     details = {
         "Topic": topic,
         "Audience": audience,