Llama 3.1 8B Code Refactor

Published by @cleancode_ai · Community adapter

code #refactor #code-quality #python #clean-code

Trained on before/after refactoring pairs extracted from open-source PRs. Suggests Extract Method, Replace Conditional with Polymorphism, and reduces cyclomatic complexity.

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("meta-llama/Meta-Llama-3.1-8B-Instruct")
base = AutoModelForCausalLM.from_pretrained("meta-llama/Meta-Llama-3.1-8B-Instruct")
model = PeftModel.from_pretrained(base, "modelforgelab/llama31-8b-refactor-lora")

inputs = tokenizer("Your prompt here", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Compatibility

transformers>=4.43
vLLM

code public

Llama 3.1 8B Bash Scripting

meta-llama/Meta-Llama-3.1-8B-Instruct