Update app.py

app.py

@@ -1,33 +1,98 @@
 import streamlit as st
 import os
 from langchain_community.chat_models import ChatHuggingFace
-from langchain_community.llms import HuggingFaceHub
-from langchain_core.messages import HumanMessage, SystemMessage
+from langchain_community.llms import HuggingFaceEndpoint
+from langchain_core.messages import HumanMessage, SystemMessage, AIMessage
 
-# ✅ Load your secret token from Hugging Face Space secrets
-hf_token = os.getenv("Data_science")  # Make sure "Data_science" is set in Space secrets
+# --- Hugging Face token setup ---
+hf_token = os.getenv("Data_science")  
 os.environ["HUGGINGFACEHUB_API_KEY"] = hf_token
 os.environ["HF_TOKEN"] = hf_token
 
-# ✅ Initialize the LLM with your token
-llm = HuggingFaceHub(
+# --- Model initialization ---
+model = HuggingFaceEndpoint(
     repo_id="meta-llama/Llama-3.2-3B-Instruct",
-    huggingfacehub_api_token=hf_token,  # <-- use hf_token here
-    model_kwargs={"temperature": 0.6, "max_new_tokens": 100}
+    provider="nebius",
+    temperature=0.6,
+    max_new_tokens=300,
+    task="conversational"
 )
 
-chat_model = ChatHuggingFace(llm=llm)
+chat_model = ChatHuggingFace(
+    llm=model,
+    repo_id="meta-llama/Llama-3.2-3B-Instruct",
+    provider="nebius",
+    temperature=0.6,
+    max_new_tokens=300,
+    task="conversational"
+)
+
+# --- Streamlit styles ---
+st.markdown("""
+<style>
+    .subject-btn {
+        display: inline-block;
+        margin: 0.3em;
+    }
+    .output-box {
+        background-color: #f9f9f9;
+        border-radius: 10px;
+        padding: 20px;
+        margin-top: 20px;
+        box-shadow: 0 2px 6px rgba(0, 0, 0, 0.05);
+    }
+</style>
+""", unsafe_allow_html=True)
+
+# --- Session state ---
+if "message_history" not in st.session_state:
+    st.session_state.message_history = []
+
+if "selected_subject" not in st.session_state:
+    st.session_state.selected_subject = "Python"
+
+# --- UI Header ---
+st.title("🎓 Data Science Mentor")
+st.markdown("Ask subject-specific questions and get guidance based on your experience level.")
+
+# --- Experience level ---
+experience = st.selectbox("👤 Select your experience level:", ["Beginner", "Intermediate", "Expert"])
+
+# --- Subject buttons ---
+st.markdown("### 📚 Choose a Subject:")
+cols = st.columns(4)
+subjects = ["Python", "SQL", "Power BI", "Statistics", "Machine Learning", "Deep Learning", "Generative AI"]
+for i, subject in enumerate(subjects):
+    if cols[i % 4].button(subject):
+        st.session_state.selected_subject = subject
+        st.session_state.message_history = []  # Reset chat on subject change
+
+# --- System Message ---
+system_prompt = f"""
+You are a highly knowledgeable data science mentor specialized in {st.session_state.selected_subject}.
+Your job is to guide a {experience.lower()} learner with clear, concise, and actionable advice.
+Explain concepts, best practices, and answer questions with patience and professionalism.
+If relevant, include example code, use-cases, or tips.
+"""
+if not st.session_state.message_history:
+    st.session_state.message_history.append(SystemMessage(content=system_prompt))
 
-# ✅ Streamlit UI
-st.title("🧠 LLaMA Chat (Data Science Mentor)")
+# --- Chat Input ---
+user_question = st.text_input(f"💬 Ask your {st.session_state.selected_subject} question:")
 
-question = st.text_input("Ask any question about data science topics:")
+if st.button("Ask Mentor"):
+    if user_question.strip():
+        with st.spinner("Thinking..."):
+            st.session_state.message_history.append(HumanMessage(content=user_question))
+            try:
+                response = chat_model.invoke(st.session_state.message_history)
+                st.session_state.message_history.append(AIMessage(content=response.content))
 
-if st.button("Ask") and question.strip():
-    messages = [
-        SystemMessage(content="You are a data science mentor."),
-        HumanMessage(content=question)
-    ]
-    response = chat_model.invoke(messages)
-    st.write("### Mentor's Response:")
-    st.write(response.content)
+                st.markdown('<div class="output-box">', unsafe_allow_html=True)
+                st.markdown("### 🧠 Mentor's Response:")
+                st.markdown(response.content)
+                st.markdown("</div>", unsafe_allow_html=True)
+            except Exception as e:
+                st.error(f"❌ Error: {e}")
+    else:
+        st.warning("⚠️ Please enter a question before submitting.")