LFFormer Evaluation

This repository contains the implementation code for the paper LF-Former: A Multi-Agent Policy with Ordered Decision Structures for Multi-Arm Manipulation (RA-Letters, 2026).

The environment definition is included in conda_env.yaml.

cd /path/to/LF-Former
conda activate lfformer

python eval.py \
  --ckpt_dir /path/to/your/checkpoint_dir \
  --ckpt_name policy_last.ckpt \
  --device auto \
  --num_rollouts 50