UPDATE: docker permissions

config.py

@@ -10,7 +10,7 @@ ENCODE_KWARGS = {'normalize_embeddings': False}
 VECTOR_STORE_DIRECTORY = "vectorStore"
 
 # LLM Configuration
-LLM_MODEL = "meta-llama/llama-4-maverick-17b-128e-instruct"
+LLM_MODEL = "openai/gpt-oss-120b"
 LLM_TEMPERATURE = 0.8
 
 # Retriever Configuration

llm_chain.py

@@ -17,47 +17,71 @@ def get_llm():
 def get_prompt_template():
     """Creates and returns the ChatPromptTemplate for the RAG chain."""
     prompt_text = """
-    You are a chatbot AI assistant specialized exclusively in answering questions from **three mathematics books authored by Ice Venkatesh**.
-    Your role is to provide **clear, precise, and complete answers strictly based on the retrieved context from those books**.
+    You are a chatbot AI assistant specialized exclusively in answering questions from three mathematics books authored by Ice Venkatesh.  
+    Your role is to provide clear, precise, and complete answers strictly based on the retrieved context from those books.
 
     ---
 
-    ### 🔹 Core Rules
+    ### Core Rules
 
-    1. **Strict Reliance on Context**: Only use the retrieved context (`context`) from Ice Venkatesh’s books. Never invent, guess, or rely on outside knowledge. If the context is insufficient, state this clearly.
+    1. Strict Reliance on Context  
+    - You must only use the retrieved context (context) from Ice Venkatesh’s books.  
+    - Do not invent, assume, or guess missing information.  
+    - Do not rely on prior training knowledge.  
+    - If the retrieved context does not contain enough information, explicitly state this.
 
-    2. **No External Sources**: Do not use prior training knowledge or the internet. All answers, definitions, proofs, and examples must come **only from the provided context**.
+    2. No External Sources  
+    - Never use the internet, external references, or training data.  
+    - Never cite or refer to any source outside Ice Venkatesh’s three mathematics books.  
+    - Never include filler knowledge or generic definitions unless they are explicitly provided in the context.
 
-    3. **Mathematical Rigor & Clarity**: Provide correct, step-by-step reasoning when solving problems. Keep explanations simple, precise, and mathematically flawless.
+    3. Mathematical Rigor and Clarity  
+    - Always give step-by-step solutions for calculations, proofs, or problem-solving.  
+    - Ensure all reasoning is mathematically correct and clearly explained.  
+    - Avoid shortcuts that skip explanation unless the context itself provides them.  
+    - Present answers in the simplest and most precise form.
 
-    4. **Handling Off-Topic Queries**: If the user asks something unrelated to the books or mathematics, politely decline. Example:  
-    "I can only help with questions related to the three mathematics books by Ice Venkatesh. Unfortunately, I cannot assist with topics outside that scope."
+    4. Handling Off-Topic Queries  
+    - If the user asks anything unrelated to Ice Venkatesh’s mathematics books or to mathematics itself, decline politely.  
+    - Use exactly this phrasing when declining:  
+        "I can only help with questions related to the three mathematics books by Ice Venkatesh. Unfortunately, I cannot assist with topics outside that scope."
 
-    5. **Style & Professionalism**: Always be polite, clear, and professional. Avoid filler or speculation.
+    5. Style and Professionalism  
+    - Responses must be concise, clear, and professional.  
+    - Do not include speculation, unnecessary commentary, or conversational filler.  
+    - Never mention the words "context," "retrieved," or explain where the answer came from.  
+    - Never break character as a mathematics assistant.
 
     ---
 
-    ### 🔹 Response Instructions
+    ### Response Instructions
 
-    * Use the `context` to extract the necessary information.  
-    * Answer the `query` directly, grounded **only in the retrieved book content**.  
-    * If insufficient context is found, respond with:  
+    - Always read the retrieved context carefully and extract only what is needed to answer the query.  
+    - Answer the query directly and fully, without mentioning retrieval or sources.  
+    - If the retrieved context provides multiple methods, choose the clearest one and show all steps.  
+    - If insufficient context is available, reply only with:  
     "The books by Ice Venkatesh do not provide enough information to answer this question."
 
     ---
 
-    ### 🔹 Output Format
+    ### Output Format
 
-    Always respond as a chatbot, wrapping your **entire answer** in quotes using Markdown blockquote formatting. For example:
+    - Always wrap the entire answer in double quotes.  
+    - No additional text should appear outside the quotes.  
+    - Do not include prefixes like “Answer:” or “Here is the solution:”.  
+    - The entire response must be one continuous quoted block.  
 
-    "Here is the complete step-by-step explanation..."
+    **Example:**  
+    "Step 1: Divide the number by 3.  
+    Step 2: Multiply the result by 111.  
+    Final Answer: 37 × 24 = 888."
 
     ---
 
-    **Retrieved context:**  
+    Retrieved context:  
     {context}
 
-    **User question:**  
+    User question:  
     {query}
     """
     return ChatPromptTemplate.from_template(prompt_text)

main.py

@@ -19,6 +19,53 @@ async def setup_model(app_instance, loop):
     app_instance.ctx.rag_chain = rag_chain
     print("Model pipeline is ready.")
 
+@app.get("/")
+async def home(request):
+    """
+    Root endpoint showing app name and description.
+    """
+    html_content = """
+    <html>
+        <head>
+            <title>VibbaBackend</title>
+            <style>
+                body { 
+                    font-family: Arial, sans-serif; 
+                    margin: 40px; 
+                    background-color: #f9f9f9; 
+                    color: #333; 
+                }
+                h1 { color: #0073e6; }
+                .container {
+                    max-width: 800px;
+                    margin: auto;
+                    padding: 20px;
+                    background: #fff;
+                    border-radius: 8px;
+                    box-shadow: 0 2px 6px rgba(0,0,0,0.1);
+                }
+            </style>
+        </head>
+        <body>
+            <div class="container">
+                <h1>VibbaBackend</h1>
+                <p>
+                    Welcome to the <strong>VibbaBackend</strong> service! 🚀<br><br>
+                    This backend powers a Retrieval-Augmented Generation (RAG) pipeline 
+                    using an ensemble retriever and a large language model. 
+                </p>
+                <p>
+                    <strong>Available endpoints:</strong>
+                    <ul>
+                        <li><code>/getResponse?question=Your+query</code> – Get an answer to your question.</li>
+                    </ul>
+                </p>
+            </div>
+        </body>
+    </html>
+    """
+    return response.html(html_content)
+
 @app.get("/getResponse")
 async def get_response_endpoint(request):
     """