Update app.py

app.py

@@ -5,28 +5,34 @@ import streamlit as st
 from sentence_transformers import SentenceTransformer
 from groq import Groq
 
-# Load FAISS index
+# ✅ Sample dataset (Replace with actual dataset later)
+data = [
+    "Alkaline electrolysis efficiency is 65-70%.",
+    "PEM electrolysis is 75-80% efficient but expensive.",
+    "SOEC electrolysis reaches 85% efficiency with heat.",
+    "Green hydrogen production costs depend on electricity price.",
+    "Solar-powered electrolysis is cost-effective in sunny regions.",
+    "Hydrogen storage is critical for economic viability.",
+]
+
+# Load FAISS index or create if missing
 index_path = "faiss_index.idx"
 model = SentenceTransformer('all-MiniLM-L6-v2')
 
-# Load FAISS index or create if missing
 if os.path.exists(index_path):
     index = faiss.read_index(index_path)
 else:
-    # Sample dataset (replace with real dataset)
-    data = [
-        "Alkaline electrolysis efficiency is 65-70%.",
-        "PEM electrolysis is 75-80% efficient but expensive.",
-        "SOEC electrolysis reaches 85% efficiency with heat.",
-    ]
     embeddings = model.encode(data).astype('float32')
     dimension = embeddings.shape[1]
     index = faiss.IndexFlatL2(dimension)
     index.add(embeddings)
     faiss.write_index(index, index_path)
 
-# Initialize Groq API
-client = Groq(api_key=os.getenv("Groq_API_Key"))
+# ✅ Set Groq API key as an environment variable 
+os.environ["GROQ_API_KEY"] = "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"  # Replace with your actual key
+
+# Initialize Groq API using the environment variable
+client = Groq(api_key=os.getenv("GROQ_API_KEY"))
 
 def rag_pipeline(query):
     """Retrieves relevant information from FAISS and queries Llama 3 via Groq API."""
@@ -37,11 +43,13 @@ def rag_pipeline(query):
     k = 3  # Retrieve top 3 relevant results
     _, indices = index.search(query_embedding, k)
     
-    # Retrieve relevant data
-    retrieved_texts = [data[i] for i in indices[0] if i < len(data)]
+    # ✅ Ensure retrieved indices are valid
+    retrieved_texts = [data[i] for i in indices[0] if 0 <= i < len(data)]
+
+    # Merge retrieved context
     context = "\n".join(retrieved_texts)
 
-    # Generate response with Llama 3
+    # ✅ Generate response with Llama 3
     response = client.chat.completions.create(
         model="llama3-8b-32k",
         messages=[
@@ -50,10 +58,9 @@ def rag_pipeline(query):
         ]
     )
 
-    # ✅ Correct way to extract the response content
     return response.choices[0].message.content  # ✅ Fix applied
 
-# Streamlit UI
+# ✅ Streamlit UI
 st.title("Green Hydrogen Techno-Economic Analyzer")
 
 query = st.text_input("Enter your question about hydrogen production:")