AI Code Generation: GitHub Copilot, CodeLlama, and Beyond

AI code generation tools boost developer productivity by suggesting code completions, generating functions, and debugging errors.

GitHub Copilot Setup

# Install Copilot extension in VS Code
# Then use inline suggestions while coding

def fibonacci(n):
    # Copilot suggests implementation
    if n <= 1:
        return n
    return fibonacci(n-1) + fibonacci(n-2)

# Copilot can generate entire functions from comments
# Type: "# function to reverse a string"
# Copilot suggests complete implementation

CodeLlama for Code Generation

from transformers import AutoTokenizer, AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained("codellama/CodeLlama-7b-hf")
tokenizer = AutoTokenizer.from_pretrained("codellama/CodeLlama-7b-hf")

prompt = "def binary_search(arr, target):"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_length=200)
code = tokenizer.decode(outputs[0])
print(code)

OpenAI for Code Generation

import openai

response = openai.chat.completions.create(
    model="gpt-4",
    messages=[
        {"role": "system", "content": "You are a Python coding expert."},
        {"role": "user", "content": "Write a function to find prime numbers up to n"}
    ]
)

generated_code = response.choices[0].message.content
print(generated_code)

Code Review with AI

code_to_review = """
def calculate_total(items):
    total = 0
    for i in range(len(items)):
        total = total + items[i]
    return total
"""

prompt = f"""
Review this code for:
1. Performance issues
2. Pythonic improvements  
3. Edge cases

Code:
{code_to_review}
"""

response = openai.chat.completions.create(
    model="gpt-4",
    messages=[{"role": "user", "content": prompt}]
)

print(response.choices[0].message.content)

AI code generation accelerates development while maintaining code quality!