metadata
license: mit
pipeline_tag: text-generation
library_name: transformers
SK²Decompile: Structure Recovery Model
This repository contains the sk2decompile-struct-6.7b model, which is part of the SK²Decompile framework introduced in the paper Decompile-Bench: Million-Scale Binary-Source Function Pairs for Real-World Binary Decompilation.
The full codebase and more details can be found on the GitHub repository.
SK²Decompile is a novel two-phase framework for binary decompilation using Large Language Models (LLMs). Our approach decomposes the complex decompilation task into two manageable phases:
Phase 1 Structure Recovery (Skeleton): Transform binary/pseudo-code into obfuscated intermediate representations (current model)
Phase 2 Identifier Naming (Skin): Generate human-readable source code with meaningful identifiers 🤗 HF Link
Usage:
python normalize_pseudo.py --input_json reverse_sample.json --output_json reverse_sample.json
python sk2decompile.py --dataset_path reverse_sample.json --model_path LLM4Binary/sk2decompile-struct-6.7b --recover_model_path LLM4Binary/sk2decompile-ident-6.7