Update app.py

app.py

@@ -2,11 +2,12 @@
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import requests
 import gradio as gr
+import torch
 
-# Load the Hugging Face model and tokenizer
-model_name = "SakanaAI/Llama-3-8B-Instruct-Coding-Expert"
+# 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)
+model = AutoModelForCausalLM.from_pretrained(model_name, load_in_8bit=True, device_map="auto")
 
 # Groq API configuration
 GROQ_API_KEY = "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"
@@ -31,8 +32,8 @@ def generate_smart_contract(language, requirements):
     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")
-    outputs = model.generate(**inputs, max_length=500)
+    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