Update app.py

app.py

@@ -10,272 +10,98 @@ from langchain.chains import RetrievalQA
 from langchain.document_loaders import PyPDFLoader
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain.chains import LLMChain
-
-# Set persistent storage path
-PERSISTENT_DIR = "vector_db"
-
-def initialize_vector_db():
-    # Check if vector database already exists in persistent storage
-    if os.path.exists(PERSISTENT_DIR) and os.listdir(PERSISTENT_DIR):
-        embeddings = HuggingFaceEmbeddings()
-        vector_db = Chroma(persist_directory=PERSISTENT_DIR, embedding_function=embeddings)
-        return None, vector_db
-    
-    base_dir = os.path.dirname(os.path.abspath(__file__))
-    pdf_files = [f for f in os.listdir(base_dir) if f.endswith('.pdf')]
-    loaders = [PyPDFLoader(os.path.join(base_dir, fn)) for fn in pdf_files]
-    
-    documents = []
-    for loader in loaders:
-        documents.extend(loader.load())
-    
-    text_splitter = RecursiveCharacterTextSplitter(
-        chunk_size=1000,
-        chunk_overlap=200,
-        length_function=len,
-        separators=["\n\n", "\n", " ", ""]
-    )
-    texts = text_splitter.split_documents(documents)
-    
-    embeddings = HuggingFaceEmbeddings()
-    vector_db = Chroma.from_documents(
-        texts, 
-        embeddings, 
-        persist_directory=PERSISTENT_DIR
+from langchain.memory import ConversationBufferMemory
+
+# Initialize session state for chat history and memory
+if "chat_history" not in st.session_state:
+    st.session_state.chat_history = []
+if "memory" not in st.session_state:
+    st.session_state.memory = ConversationBufferMemory(
+        memory_key="chat_history",
+        return_messages=True
     )
-    vector_db.persist()
-    return documents, vector_db
-
-# System instructions for the LLM
-system_prompt = """You are an expert organic farming consultant with specialization in Agro-Homeopathy. When providing suggestions and remedies:
-1. Always specify medicine potency as 6c unless the uploaded text mentions some other value explicitly
-3. Provide comprehensive diagnosis and treatment advice along with organic farming best practices applicable in the given context
-4. Base recommendations on homeopathic and organic farming principles
-"""
-
-api_key1 = os.getenv("api_key")
-
-start_time = time.time()
-st.set_page_config(page_title="Dr. Radha: The Agro-Homeopath", page_icon="🚀", layout="wide")
-
-# CSS for dark green background and white text
-st.markdown("""
-<style>
-    /* Set background color for entire app */
-    .stApp {
-        background-color: #1B4D3E !important;
-        color: white !important;
-    }
-    
-    /* Style input fields */
-    .stTextInput>div>div>input {
-        color: black !important;
-        background-color: rgba(255,255,255,0.1) !important;
-    }
-    
-    /* Style buttons */
-    .stButton>button {
-        color: black !important;
-        background-color: yellow !important;
-    }
-    
-    }
-</style>
-""", unsafe_allow_html=True)
-
-st.markdown("""
-<style>
-    #the-title {
-        text-align: center;
-        font-size: 24px;
-        color: white;
-    }
-</style>
-""", unsafe_allow_html=True)
-
-st.title("🌿 Dr. Radha: AI-Powered Organic Farming Consultant")
-st.subheader("Specializing in Agro-Homeopathy | Free Consultation")
-
-# Add information request message
-st.markdown("""
-Please provide complete details about the issue, including:
-- Detailed description of plant problem
-- Current location, temperature & weather conditions
-""")
-
-human_image = "human.png"
-robot_image = "bot.jpg"
-
-# Set up Groq API with temperature 0.7
-llm = ChatGroq(
-    api_key=api_key1, 
-    max_tokens=None, 
-    timeout=None, 
-    max_retries=2, 
-    temperature=0.7,
-    model="llama-3.1-70b-versatile"
-)
-
-embeddings = HuggingFaceEmbeddings()
-end_time = time.time()
-print(f"Setting up Groq LLM & Embeddings took {end_time - start_time:.4f} seconds")
-
-# Initialize session state
-if "documents" not in st.session_state:
-    st.session_state["documents"] = None
-if "vector_db" not in st.session_state:  
-    st.session_state["vector_db"] = None
-if "query" not in st.session_state:
-    st.session_state["query"] = ""
-    
-start_time = time.time()
-if st.session_state["documents"] is None or st.session_state["vector_db"] is None:
-    with st.spinner("Loading data..."):
-        documents, vector_db = initialize_vector_db()
-        st.session_state["documents"] = documents  
-        st.session_state["vector_db"] = vector_db
-else:
-    documents = st.session_state["documents"]
-    vector_db = st.session_state["vector_db"]
 
-end_time = time.time()
-print(f"Loading and processing PDFs & vector database took {end_time - start_time:.4f} seconds")
+# Add sidebar
+st.sidebar.title("Chat History")
 
-start_time = time.time()
-retriever = vector_db.as_retriever()
+# Add Clear Session button in sidebar
+if st.sidebar.button("Clear Session"):
+    st.session_state.chat_history = []
+    st.session_state.memory.clear()
+    st.experimental_rerun()
 
-prompt_template = """As an expert organic farming consultant with specialization in Agro-Homeopathy, analyze the following context and question to provide a clear, structured response.
+# Display chat history in sidebar
+for i, chat in enumerate(st.session_state.chat_history):
+    st.sidebar.text(f"Q{i+1}: {chat['question']}")
+    if st.sidebar.button(f"Go to Q{i+1}", key=f"btn_{i}"):
+        st.session_state.query = chat['question']
 
-Context: {context}
-
-Question: {question}
-
-Provide your response in the following format:
-Analysis: Analyze the described plant condition
-Treatment: Recommend relevant organic farming principles and specific homeopathic medicine(s) with exact potency and repetition frequency. Suggest a maximum of 4 medicines in the order of relevance for any single problem. 
-Instructions for Use:
-Small Plots or Gardens: Make sure your dispensing equipment is not contaminated with
-other chemicals or fertilisers as these may antidote the energetic effects of the treatment—
-rinse well with hot water before use if necessary. Add one pill to each 200 ml of water, shake
-vigorously, and then spray or water your plants. Avoid using other chemicals or fertilisers for
-10 days following treatment so that the energetic effects of the treatment are not antidoted.
-(One vial of 100 pills makes 20 litres. Plants remain insect or disease free for up to 3 months
-following one treatment.)
-Large Plots or Farms: Add the remedy to water and apply with the dispensing device of
-your choice: watering can, backpack sprayer, boomspray, reticulation systems (add to tanks
-or pumps). Make sure your dispensing equipment is not contaminated with other chemicals
-or fertilisers as these may antidote the energetic effects of the treatment—rinse with hot
-water or steam clean before use if necessary. Avoid using other chemicals or fertilisers for
-10 days following treatment.
-Dosage rates are approximate and may vary according to different circumstances and
-experiences. Suggested doses are:
-10 pills or 10ml/10 litre on small areas,
-500 pills or 125ml/500l per hectare,
-1000 pills or 250ml/500l per hectare,
-2500 pills or 500ml/500l per hectare,
-Add pills or liquid to your water and mix (with a stick if necessary for large containers).
-
-Recommendations: Provide couple of key pertinent recommendations based on the query
-
-Remember to maintain a professional, clear tone and ensure all medicine recommendations include specific potency.
-
-Answer:"""
-
-# Create the QA chain with correct variables
+# Update the QA chain to include memory
 qa = RetrievalQA.from_chain_type(
     llm=llm,
     chain_type="stuff",
     retriever=retriever,
+    memory=st.session_state.memory,
     chain_type_kwargs={
         "prompt": PromptTemplate(
             template=prompt_template,
-            input_variables=["context", "question"]
+            input_variables=["context", "question", "chat_history"]
         )
     }
 )
 
-# Create a separate LLMChain for fallback
-fallback_template = """As an expert organic farming consultant with specialization in Agro-Homeopathy, analyze the following context and question to provide a clear, structured response.
-
-Question: {question}
-
-Format your response as follows:
-Analysis: Analyze the described plant condition
-Treatment: Recommend relevant organic farming principles and specific homeopathic medicine(s) with exact potency and repetition frequency. Suggest a maximum of 4 medicines in the order of relevance for any single problem.
-Instructions for Use:
-Small Plots or Gardens: Make sure your dispensing equipment is not contaminated with
-other chemicals or fertilisers as these may antidote the energetic effects of the treatment—
-rinse well with hot water before use if necessary. Add one pill to each 200 ml of water, shake
-vigorously, and then spray or water your plants. Avoid using other chemicals or fertilisers for
-10 days following treatment so that the energetic effects of the treatment are not antidoted.
-(One vial of 100 pills makes 20 litres. Plants remain insect or disease free for up to 3 months
-following one treatment.)
-Large Plots or Farms: Add the remedy to water and apply with the dispensing device of
-your choice: watering can, backpack sprayer, boomspray, reticulation systems (add to tanks
-or pumps). Make sure your dispensing equipment is not contaminated with other chemicals
-or fertilisers as these may antidote the energetic effects of the treatment—rinse with hot
-water or steam clean before use if necessary. Avoid using other chemicals or fertilisers for
-10 days following treatment.
-Dosage rates are approximate and may vary according to different circumstances and
-experiences. Suggested doses are:
-10 pills or 10ml/10 litre on small areas
-500 pills or 125ml/500l per hectare
-1000 pills or 250ml/500l per hectare
-2500 pills or 500ml/500l per hectare
-Add pills or liquid to your water and mix (with a stick if necessary for large containers).
-
-Recommendations: Provide couple of key pertinent recommendations based on the query
-
-Maintain a professional tone and ensure all medicine recommendations include specific potency.
-
-Answer:"""
-
-fallback_prompt = PromptTemplate(template=fallback_template, input_variables=["question"])
-fallback_chain = LLMChain(llm=llm, prompt=fallback_prompt)
-
-chat_container = st.container()
-
-st.markdown("""
-<style>
-.stButton>button {
-    color: black !important;
-    background-color: yellow !important;
-}
-</style>
-""", unsafe_allow_html=True)
-
-with st.form(key='query_form'):
-    query = st.text_input("Ask your question:", value="")
-    submit_button = st.form_submit_button(label='Submit')
-
-end_time = time.time()
-#print(f"Setting up retrieval chain took {end_time - start_time:.4f} seconds")
-start_time = time.time()
+# Update fallback chain to include memory
+fallback_chain = LLMChain(
+    llm=llm, 
+    prompt=fallback_prompt,
+    memory=st.session_state.memory
+)
 
+# Update the form submission handling
 if submit_button and query:
     with st.spinner("Generating response..."):
-        result = qa({"query": query})  
+        # Add query to chat history
+        st.session_state.chat_history.append({
+            "question": query,
+            "timestamp": time.time()
+        })
+        
+        # Get response with memory context
+        result = qa({
+            "query": query,
+            "chat_history": st.session_state.memory.chat_memory
+        })
+        
         if result['result'].strip() == "":
-            # If no result from PDF, use fallback chain
             fallback_result = fallback_chain.run(query)
             response = fallback_result
         else:
             response = result['result']
+        
+        # Store the interaction in memory
+        st.session_state.memory.save_context(
+            {"input": query},
+            {"output": response}
+        )
+
+    # Display chat messages
+    for chat in st.session_state.chat_history:
+        col1, col2 = st.columns([1, 10])
+        with col1:
+            st.image(human_image, width=80)
+        with col2:
+            st.markdown(f"{chat['question']}")
+            
+        col1, col2 = st.columns([1, 10])
+        with col1:
+            st.image(robot_image, width=80)
+        with col2:
+            if chat['question'] == query:
+                st.markdown(f"{response}")
+            else:
+                # Retrieve previous response from memory
+                prev_response = st.session_state.memory.load_memory_variables({})
+                st.markdown(f"{prev_response}")
 
-    col1, col2 = st.columns([1, 10])
-    with col1:
-        st.image(human_image, width=80)
-    with col2:
-        st.markdown(f"{query}")
-    col1, col2 = st.columns([1, 10])
-    with col1:
-        st.image(robot_image, width=80)
-    with col2:
-        st.markdown(f"{response}")
-    
     st.markdown("---")
-    
     st.session_state["query"] = ""
-
-end_time = time.time()
-print(f"Actual query took {end_time - start_time:.4f} seconds")