Prerequisites
Install
python -m pip install -r requirements.txt
python -m pip install -e .[dev]
or pip install -r dev-requirements.txt
Model (MLX)
python -c "from mlx_gen_parity.interop import convert_hf_to_mlx; convert_hf_to_mlx('Qwen/Qwen2.5-0.5B', quantize=False, local_out='mlx_qwen2_0_5b')"
Run GSPO
make run-gspo-eca MODEL=./mlx_qwen2_0_5b
make run-gspo-life MODEL=./mlx_qwen2_0_5b
Config-driven
python scripts/run_from_config.py --config configs/rl_eca.yaml
Tests
pytest -q