pylingual/dev_scripts/README.md

Model Training

PyLingual's accuracy is dependent on having accurate segmentation and statement models ¹ . The segmentation model divides a list of bytecode instructions into groups for each source instruction. The statement model transforms each group of instructions into source code. The instructions for training these models is as follows:

Dataset generation

Create a dataset JSON file based off the sample (sample_jsons/py36-sample-data.json).

The dataset directory should be structured like so, with only one .py file per directory:

dataset
├── 0
│   └── file.py
├── 1
│   └── file.py
...
├── 999
│   └── file.py
└── 1000
    └── file.py

The names of the inner directories and files do not matter. Then create the dataset:

uv run prepare_dataset.py <path to JSON>

Segmentation model

Create a segmentation model JSON file based off the sample (sample_jsons/py36-sample-segmentation.json). Then train the model:

uv run train_models.py --segmentation <path to JSON>

Statement model

Create a statement model JSON file based off the sample (sample_jsons/py36-sample-statement.json). Then train the model:

uv run train_models.py --statement <path to JSON>

Once models are trained, update ../pylingual/decompiler_config.yaml or create a separate config file by replacing the old models with the newly trained ones.

---

1. pylingual models. ↩︎