Update app.py

app.py

@@ -3,9 +3,10 @@ import streamlit as st
 from q_learning_chatbot import QLearningChatbot
 from xgb_mental_health import MentalHealthClassifier
 from bm25_retreive_question import QuestionRetriever as QuestionRetriever_bm25
-from Chromadb_storage import QuestionRetriever as QuestionRetriever_chromaDB
+from Chromadb_storage_JyotiNigam import QuestionRetriever as QuestionRetriever_chromaDB
 from llm_response_generator import LLLResponseGenerator
 import os
+
 # Streamlit UI
 st.title("FOMO Fix - RL-based Mental Health Assistant")
 
@@ -26,7 +27,7 @@ chatbot = QLearningChatbot(states, actions)
 # data_path = "/Users/jaelinlee/Documents/projects/fomo/input/data.csv"
 data_path = "data/data.csv"
 tokenizer_model_name = "nlptown/bert-base-multilingual-uncased-sentiment"
-mental_classifier_model_path = "mental_health_model.pkl"
+mental_classifier_model_path = "app/mental_health_model.pkl"
 mental_classifier = MentalHealthClassifier(data_path, mental_classifier_model_path)
 
 
@@ -147,6 +148,10 @@ if user_message:
         ai_tone = chatbot.get_action(user_sentiment)
         print(ai_tone)
 
+        print(st.session_state.messages)
+
+
+
         # LLM Response Generator
         HUGGINGFACEHUB_API_TOKEN = os.getenv('HUGGINGFACEHUB_API_TOKEN')
 
@@ -154,18 +159,19 @@ if user_message:
         temperature = 0.1
         max_length = 128
 
+        # Collect all messages exchanged so far into a single text string
+        all_messages = "\n".join([message.get("content") for message in st.session_state.messages])
+
         #Question asked to the user: {question}
 
         template = """INSTRUCTIONS: {context}
 
             Respond to the user with a tone of {ai_tone}.
 
-            Question asked to the user: "None"
-
             Response by the user: {user_text}
             Response;
             """
-        context = "You are a mental health supporting non-medical assistant. Provide some advice and ask a relevant question back to the user."
+        context = f"You are a mental health supporting non-medical assistant. Provide some advice and ask a relevant question back to the user. {all_messages}"
 
         llm_response = llm_model.llm_inference(
             model_type="huggingface",
@@ -184,6 +190,7 @@ if user_message:
         else:
             llm_reponse_with_quesiton = llm_response
 
+        # Append the user and AI responses to the chat history
         st.session_state.messages.append({"role": "ai", "content": llm_reponse_with_quesiton})
 
     with st.chat_message("ai"):
@@ -220,7 +227,7 @@ with st.sidebar.expander('Behind the Scene', expanded=section_visible):
     st.write(f"- AI Tone: {st.session_state.ai_tone.capitalize()}")
     st.write(f"- Question retrieved from: {selected_retriever_option}")
     st.write(
-        f"- If the user feels negative or moderately negative, at the end of the AI response, it adds a mental health condition related question. The question is retrieved from DB. The categories of questions are limited to Depression, Anxiety, and ADHD which are most associated with FOMO related to excessive social media usage."
+        f"- If the user feels negative, moderately negative, or neutral, at the end of the AI response, it adds a mental health condition related question. The question is retrieved from DB. The categories of questions are limited to Depression, Anxiety, and ADHD which are most associated with FOMO related to excessive social media usage."
     )
     st.write(
         f"- Below q-table is continuously updated after each interaction with the user. If the user's mood increases, AI gets a reward. Else, AI gets a punishment."