app.py

import os
import gradio as gr
from huggingface_hub import InferenceClient

class XylariaChat:
    def __init__(self):
        # Securely load HuggingFace token
        self.hf_token = os.getenv("HF_TOKEN")
        if not self.hf_token:
            raise ValueError("HuggingFace token not found in environment variables")
        
        # Initialize the inference client
        self.client = InferenceClient(
            model= os.getenv("MODEL_NAME"), 
            api_key=self.hf_token
        )
        
        # Initialize conversation history and persistent memory
        self.conversation_history = []
        self.persistent_memory = {}
        
        # System prompt with more detailed instructions
        self.system_prompt = """You are a helpful and harmless assistant. You are Xylaria developed by Sk Md Saad Amin. You should think step-by-step. 
"""
    def store_information(self, key, value):
        """Store important information in persistent memory"""
        self.persistent_memory[key] = value

    def retrieve_information(self, key):
        """Retrieve information from persistent memory"""
        return self.persistent_memory.get(key)

    def reset_conversation(self):
        """
        Completely reset the conversation history, persistent memory, 
        and clear API-side memory
        """
        # Clear local memory
        self.conversation_history = []
        self.persistent_memory.clear()
        
        # Clear API-side memory by resetting the conversation
        try:
            # Attempt to clear any API-side session or context
            self.client = InferenceClient(
                model="Qwen/QwQ-32B-Preview", 
                api_key=self.hf_token
            )
        except Exception as e:
            print(f"Error resetting API client: {e}")
        
        return None  # To clear the chatbot interface

    def get_response(self, user_input):
        # Prepare messages with conversation context and persistent memory
        messages = [
            {"role": "system", "content": self.system_prompt},
            *self.conversation_history,
            {"role": "user", "content": user_input}
        ]
        
        # Add persistent memory context if available
        if self.persistent_memory:
            memory_context = "Remembered Information:\n" + "\n".join(
                [f"{k}: {v}" for k, v in self.persistent_memory.items()]
            )
            messages.insert(1, {"role": "system", "content": memory_context})
        
        # Generate response with streaming
        try:
            stream = self.client.chat.completions.create(
                messages=messages,
                temperature=0.5,
                max_tokens=10240,
                top_p=0.7,
                stream=True
            )
            
            return stream
        
        except Exception as e:
            return f"Error generating response: {str(e)}"

    def create_interface(self):
        def streaming_response(message, chat_history):
            # Clear input textbox
            response_stream = self.get_response(message)
            
            # If it's an error, return immediately
            if isinstance(response_stream, str):
                return "", chat_history + [[message, response_stream]]
            
            # Prepare for streaming response
            full_response = ""
            updated_history = chat_history + [[message, ""]]
            
            # Streaming output
            for chunk in response_stream:
                if chunk.choices[0].delta.content:
                    chunk_content = chunk.choices[0].delta.content
                    full_response += chunk_content
                    
                    # Update the last message in chat history with partial response
                    updated_history[-1][1] = full_response
                    yield "", updated_history
            
            # Update conversation history
            self.conversation_history.append(
                {"role": "user", "content": message}
            )
            self.conversation_history.append(
                {"role": "assistant", "content": full_response}
            )
            
            # Limit conversation history to prevent token overflow
            if len(self.conversation_history) > 10:
                self.conversation_history = self.conversation_history[-10:]

        # Custom CSS for Inter font
        custom_css = """
        @import url('https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600;700&display=swap');
        
        body, .gradio-container {
            font-family: 'Inter', sans-serif !important;
        }
        
        .chatbot-container .message {
            font-family: 'Inter', sans-serif !important;
        }
        
        .gradio-container input, 
        .gradio-container textarea, 
        .gradio-container button {
            font-family: 'Inter', sans-serif !important;
        }
        """

        with gr.Blocks(theme='soft', css=custom_css) as demo:
            # Chat interface with improved styling
            with gr.Column():
                chatbot = gr.Chatbot(
                    label="Xylaria 1.4 Senoa",
                    height=500,
                    show_copy_button=True
                )
                
                # Input row with improved layout
                with gr.Row():
                    txt = gr.Textbox(
                        show_label=False, 
                        placeholder="Type your message...", 
                        container=False,
                        scale=4
                    )
                    btn = gr.Button("Send", scale=1)
                
                # Clear history and memory buttons
                clear = gr.Button("Clear Conversation")
                clear_memory = gr.Button("Clear Memory")
            
            # Submit functionality with streaming
            btn.click(
                fn=streaming_response, 
                inputs=[txt, chatbot], 
                outputs=[txt, chatbot]
            )
            txt.submit(
                fn=streaming_response, 
                inputs=[txt, chatbot], 
                outputs=[txt, chatbot]
            )
            
            # Clear conversation history
            clear.click(
                fn=lambda: None, 
                inputs=None, 
                outputs=[chatbot],
                queue=False
            )
            
            # Clear persistent memory and reset conversation
            clear_memory.click(
                fn=self.reset_conversation,
                inputs=None,
                outputs=[chatbot],
                queue=False
            )

            # Ensure memory is cleared when the interface is closed
            demo.load(self.reset_conversation, None, None)
        
        return demo

# Launch the interface
def main():
    chat = XylariaChat()
    interface = chat.create_interface()
    interface.launch(
        share=True,  # Optional: create a public link
        debug=True   # Show detailed errors
    )

if __name__ == "__main__":
    main()