app.py

import gradio as gr
from huggingface_hub import InferenceClient
import os

HF_TOKEN = os.getenv("HF_TOKEN")

# Utilizing the conversational task through chat_completion
client = InferenceClient("Qwen/Qwen2.5-7B-Instruct", token=HF_TOKEN)

class StateController:
    def __init__(self):
        self.state_array = [0] * 121
        self.base_metric = 60
        self.batch_unit = 10
        self.memory_register = {}

    def initialize_grid(self):
        for i in range(51):
            self.state_array[i] = {"Blocks": i // self.batch_unit, "Units": i % self.batch_unit}
        return "Grid initialized: 5 active blocks."

    def render_grid(self):
        grid_output = ""
        for i in range(121):
            if i == 120:
                grid_output += " [NODE_120] "
            elif i % 10 == 0:
                grid_output += "<"
            else:
                grid_output += "."
        return grid_output

    def resolve_grid(self):
        self.memory_register["STATUS"] = "RESOLVED"
        self.state_array = [0] * 121
        return "System resolved. State array reset to zero."

def generate_response(message, history):
    # Hardware diagnostic override
    if "run grid diagnostic" in message.lower():
        controller = StateController()
        output = "Diagnostic sequence initiated.\n\n"
        output += f"{controller.initialize_grid()}\n\n"
        output += "Rendering 121-point array:\n"
        output += f"{controller.render_grid()}\n\n"
        output += "Executing state resolution:\n"
        output += f"{controller.resolve_grid()}"
        return output

    system_instruction = (
        "You are a logic-focused inference engine. "
        "You utilize strict state-hold memory and parallel integer blocks. "
        "Provide direct, technical, and accurate responses."
    )
    
    # Correct format for conversational task
    messages = [{"role": "system", "content": system_instruction}]
    for user_msg, assistant_msg in history:
        messages.append({"role": "user", "content": user_msg})
        messages.append({"role": "assistant", "content": assistant_msg})
    messages.append({"role": "user", "content": message})

    try:
        # Switching to chat_completion for model compatibility
        response = client.chat_completion(
            messages,
            max_tokens=1024,
            stream=False
        )
        return response.choices[0].message.content
    except Exception as error:
        return f"System Error: {str(error)}. Verify your token permissions."

custom_css = """
body, .gradio-container { background-color: #0b0f19 !important; }
footer {display: none !important}
.message.user { background-color: #1e293b !important; border: 1px solid #3b82f6 !important; }
.message.bot { background-color: #0f172a !important; color: #60a5fa !important; }
"""

with gr.Blocks(theme=gr.themes.Soft(primary_hue="blue"), css=custom_css) as demo:
    gr.Markdown("# Advanced Logic Interface")
    gr.ChatInterface(
        fn=generate_response,
        description="Inference layer utilizing state-hold logic.",
        examples=[
            "Run grid diagnostic",
            "Calculate the integer distribution for 120 units across 3 nodes.",
            "Explain network latency using technical terminology."
        ]
    )

if __name__ == "__main__":
    demo.launch()