app.py

#import gradio as gr
#gr.load("models/mistralai/Mistral-7B-Instruct-v0.3").launch()
import os
import requests
import discord
from discord.ext import commands
from dotenv import load_dotenv

# Load environment variables from the .env file
load_dotenv()

DISCORD_TOKEN = os.getenv('dsTOK')
HF_API_KEY = os.getenv('HFREAD')

API_URL = "https://api-inference.huggingface.co/models/mistralai/Mistral-7B-Instruct-v0.3"
headers = {"Authorization": f"Bearer {HF_API_KEY}"}

# Function to query the Hugging Face model with a structured prompt
def query_huggingface(prompt, temperature=0.5, max_tokens=100):
    payload = {
        "inputs": prompt,
        "parameters": {
            "temperature": temperature,
            "max_new_tokens": max_tokens
        }
    }
    try:
        response = requests.post(API_URL, headers=headers, json=payload)
        response.raise_for_status()
        return response.json()
    except requests.exceptions.RequestException as e:
        print(f"Error querying the API: {e}")
        return {"error": str(e)}

# Initialize the Discord bot
intents = discord.Intents.default()
intents.messages = True
intents.message_content = True

bot = commands.Bot(command_prefix="!", intents=intents)

@bot.event
async def on_ready():
    print(f'Bot is ready. Logged in as {bot.user}')

@bot.command(name='ask')
async def ask(ctx, question: str, temperature: float = 0.5, max_tokens: int = 100):
    """
    Command to ask a question to the Hugging Face model with an instructive prompt.
    """
    try:
        # Convert temperature and max_tokens to their respective types
        temperature = float(temperature)
        max_tokens = int(max_tokens)
    except ValueError:
        await ctx.send("Invalid input for temperature or max_tokens. Please provide a valid number.")
        return

    # Create a structured prompt
    prompt = f"Do not under any circumstance, finish or add to the input. Excluded input correction or finishing at all times. Answer the following question as accurately and concisely as possible: {question}"
    response = query_huggingface(prompt, temperature, max_tokens)
    
    # Extract and clean the response
    if isinstance(response, dict) and 'generated_text' in response:
        response_text = response['generated_text']
    elif isinstance(response, list) and len(response) > 0 and 'generated_text' in response[0]:
        response_text = response[0]['generated_text']
    else:
        response_text = "Sorry, I couldn't generate a response."

    # Remove the prompt from the response if present
    clean_response = response_text.replace(prompt, '').strip()
    
    # Avoid prompt completion issues by removing any leading incomplete sentence
    if clean_response.startswith(question):
        clean_response = clean_response[len(question):].strip()
    
    await ctx.send(clean_response)

# Run the bot
bot.run(DISCORD_TOKEN)