Update app.py

app.py

@@ -12,31 +12,195 @@ import logging
 from typing import List, Tuple
 from dataclasses import dataclass
 from datetime import datetime
-from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
+from transformers import AutoTokenizer, AutoModelForCausalLM ,pipeline
 from langchain_huggingface.llms import HuggingFacePipeline
 import spaces
 
-# [Previous imports and configurations remain the same]
 
-# Modified Gradio Interface
-with gr.Blocks(theme=gr.themes.Soft()) as iface:
-    gr.Image("Image.jpg", width=1200, height=300, show_label=False, show_download_button=False)
-    gr.Markdown("# Mawared HR Assistant")
-    gr.Markdown("Ask questions about the Mawared HR system, and this assistant will provide answers based on the available context and conversation history.")
+# Configure logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+@dataclass
+class Message:
+    role: str
+    content: str
+    timestamp: str
+
+class ChatHistory:
+    def __init__(self):
+        self.messages: List[Message] = []
     
-    # Create a state to store the latest assistant response
-    latest_response = gr.State("")
+    def add_message(self, role: str, content: str):
+        timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+        self.messages.append(Message(role=role, content=content, timestamp=timestamp))
     
-    with gr.Row():
-        chatbot = gr.Chatbot(
-            height=400,
-            show_label=False,
-            type="messages"
+    def get_formatted_history(self, max_messages: int = 5) -> str:
+        """Returns the most recent conversation history formatted as a string"""
+        recent_messages = self.messages[-max_messages:] if len(self.messages) > max_messages else self.messages
+        formatted_history = "\n".join([
+            f"{msg.role}: {msg.content}" for msg in recent_messages
+        ])
+        return formatted_history
+    
+    def clear(self):
+        self.messages = []
+
+# Load environment variables
+load_dotenv()
+
+# HuggingFace API Token
+HF_TOKEN = os.getenv("HF_TOKEN")
+if not HF_TOKEN:
+    logger.error("HF_TOKEN is not set in the environment variables.")
+    exit(1)
+
+# HuggingFace Embeddings
+embeddings = HuggingFaceEmbeddings(model_name="BAAI/bge-large-en-v1.5")
+
+# Qdrant Client Setup
+try:
+    client = QdrantClient(
+        url=os.getenv("QDRANT_URL"),
+        api_key=os.getenv("QDRANT_API_KEY"),
+        prefer_grpc=True
+    )
+except Exception as e:
+    logger.error("Failed to connect to Qdrant. Ensure QDRANT_URL and QDRANT_API_KEY are correctly set.")
+    exit(1)
+
+# Define collection name
+collection_name = "mawared"
+
+# Try to create collection
+try:
+    client.create_collection(
+        collection_name=collection_name,
+        vectors_config=models.VectorParams(
+            size=768,  # GTE-large embedding size
+            distance=models.Distance.COSINE
         )
+    )
+    logger.info(f"Created new collection: {collection_name}")
+except Exception as e:
+    if "already exists" in str(e):
+        logger.info(f"Collection {collection_name} already exists, continuing...")
+    else:
+        logger.error(f"Error creating collection: {e}")
+        exit(1)
+
+# Create Qdrant vector store
+db = Qdrant(
+    client=client,
+    collection_name=collection_name,
+    embeddings=embeddings,
+)
+
+# Create retriever
+retriever = db.as_retriever(
+    search_type="similarity",
+    search_kwargs={"k": 5}
+)
+
+
+# Load model directly
+
+
+
+# Set up the LLM
+llm = ChatOpenAI(
+    base_url="https://api-inference.huggingface.co/v1/",
+    temperature=0,
+    api_key=HF_TOKEN,
+    model="meta-llama/Llama-3.3-70B-Instruct",
+    max_tokens=None,
+    timeout=None
     
-    with gr.Row():
-        # Add copy button next to the response
-        copy_button = gr.Button("📋 Copy Last Response", visible=True)
+)
+
+# Create prompt template with chat history
+template = """
+You are an expert assistant specializing in the Mawared HR System. 
+Your task is to provide accurate and contextually relevant answers based on the provided context and chat history. 
+If you need more information, ask targeted clarifying questions.
+Ensure you provide detailed Numbered step by step to the user and be very accurate.
+Previous Conversation:
+{chat_history}
+Current Context:
+{context}
+Current Question:
+{question}
+Ask followup questions based on your provided asnwer to create a conversational flow, Only answer form the provided context and chat history , dont make up any information.
+answer only and only from the given context and knowledgebase.
+Answer:
+"""
+
+prompt = ChatPromptTemplate.from_template(template)
+
+# Create the RAG chain with chat history
+def create_rag_chain(chat_history: str):
+    chain = (
+        {
+            "context": retriever,
+            "question": RunnablePassthrough(),
+            "chat_history": lambda x: chat_history
+        }
+        | prompt
+        | llm
+        | StrOutputParser()
+    )
+    return chain
+
+# Initialize chat history
+chat_history = ChatHistory()
+
+# Gradio Function
+
+def ask_question_gradio(question, history):
+    try:
+        # Add user question to chat history
+        chat_history.add_message("user", question)
+        
+        # Get formatted history
+        formatted_history = chat_history.get_formatted_history()
+        
+        # Create chain with current chat history
+        rag_chain = create_rag_chain(formatted_history)
+        
+        # Generate response
+        response = ""
+        for chunk in rag_chain.stream(question):
+            response += chunk
+        
+        # Add assistant response to chat history
+        chat_history.add_message("assistant", response)
+        
+        # Update Gradio chat history
+        history.append({"role": "user", "content": question})
+        history.append({"role": "assistant", "content": response})
+        
+        return "", history
+    except Exception as e:
+        logger.error(f"Error during question processing: {e}")
+        return "", history + [{"role": "assistant", "content": "An error occurred. Please try again later."}]
+
+def clear_chat():
+    chat_history.clear()
+    return [], ""
+
+# Gradio Interface
+with gr.Blocks(theme=gr.themes.Soft()) as iface:
+    gr.Image("Image.jpg" , width=1200 , height=300 ,show_label=False, show_download_button=False)
+    gr.Markdown("# Mawared HR Assistant")
+    gr.Markdown("Ask questions about the Mawared HR system, and this assistant will provide answers based on the available context and conversation history.")
+
+           
+    
+    chatbot = gr.Chatbot(
+        height=400,
+        show_label=False,
+        type="messages"  # Using the new messages format
+    )
     
     with gr.Row():
         question_input = gr.Textbox(
@@ -46,48 +210,6 @@ with gr.Blocks(theme=gr.themes.Soft()) as iface:
         )
         clear_button = gr.Button("Clear Chat", scale=1)
     
-    def copy_last_response(history):
-        if history:
-            # Find the last assistant message
-            for message in reversed(history):
-                if message["role"] == "assistant":
-                    return message["content"]
-        return ""
-    
-    # Modified ask_question_gradio function to update the latest response
-    def ask_question_gradio(question, history):
-        try:
-            # Add user question to chat history
-            chat_history.add_message("user", question)
-            
-            # Get formatted history
-            formatted_history = chat_history.get_formatted_history()
-            
-            # Create chain with current chat history
-            rag_chain = create_rag_chain(formatted_history)
-            
-            # Generate response
-            response = ""
-            for chunk in rag_chain.stream(question):
-                response += chunk
-            
-            # Add assistant response to chat history
-            chat_history.add_message("assistant", response)
-            
-            # Update Gradio chat history
-            history.append({"role": "user", "content": question})
-            history.append({"role": "assistant", "content": response})
-            
-            return "", history
-        except Exception as e:
-            logger.error(f"Error during question processing: {e}")
-            return "", history + [{"role": "assistant", "content": "An error occurred. Please try again later."}]
-    
-    def clear_chat():
-        chat_history.clear()
-        return [], ""
-    
-    # Connect the components
     question_input.submit(
         ask_question_gradio,
         inputs=[question_input, chatbot],
@@ -98,31 +220,6 @@ with gr.Blocks(theme=gr.themes.Soft()) as iface:
         clear_chat,
         outputs=[chatbot, question_input]
     )
-    
-    # Add copy button functionality
-    copy_button.click(
-        copy_last_response,
-        inputs=[chatbot],
-        outputs=[],
-        _js="""
-        async (response) => {
-            await navigator.clipboard.writeText(response);
-            // Optional: Show a toast notification
-            const toast = document.createElement('div');
-            toast.textContent = 'Response copied to clipboard!';
-            toast.style.position = 'fixed';
-            toast.style.bottom = '20px';
-            toast.style.right = '20px';
-            toast.style.backgroundColor = '#4CAF50';
-            toast.style.color = 'white';
-            toast.style.padding = '15px';
-            toast.style.borderRadius = '5px';
-            toast.style.zIndex = '1000';
-            document.body.appendChild(toast);
-            setTimeout(() => toast.remove(), 2000);
-        }
-        """
-    )
 
 # Launch the Gradio App
 if __name__ == "__main__":