app.py

# app.py
from transformers import AutoTokenizer, AutoModelForCausalLM
import requests
import gradio as gr
import torch

# Load the Hugging Face model and tokenizer in 8-bit precision
model_name = "gpt2"  # Smaller and faster model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, load_in_8bit=True, device_map="auto")

# Groq API configuration
GROQ_API_KEY = "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"
GROQ_API_URL = "https://api.groq.com/v1/completions"

# Function to query Groq API
def query_groq(prompt):
    headers = {
        "Authorization": f"Bearer {GROQ_API_KEY}",
        "Content-Type": "application/json"
    }
    data = {
        "prompt": prompt,
        "max_tokens": 150
    }
    response = requests.post(GROQ_API_URL, headers=headers, json=data)
    return response.json()["choices"][0]["text"]

# Function to generate smart contract code
def generate_smart_contract(language, requirements):
    # Create a prompt for the model
    prompt = f"Generate a {language} smart contract with the following requirements: {requirements}"
    
    # Use the Hugging Face model to generate code
    inputs = tokenizer(prompt, return_tensors="pt").to("cuda")  # Move inputs to GPU
    outputs = model.generate(**inputs, max_length=300)  # Reduced max_length
    generated_code = tokenizer.decode(outputs[0], skip_special_tokens=True)
    
    # Enhance the code using Groq API
    enhanced_code = query_groq(generated_code)
    
    return enhanced_code

# Gradio interface for the app
def generate_contract(language, requirements):
    return generate_smart_contract(language, requirements)

interface = gr.Interface(
    fn=generate_contract,
    inputs=["text", "text"],
    outputs="text",
    title="Smart Contract Generator",
    description="Generate smart contracts using AI."
)

# Launch the Gradio app
if __name__ == "__main__":
    interface.launch()