This is the PyTorch implementation of Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware. Code in this repository is copied from the official codebase and simplified such that minimal requirements are demanded. Also, sufficient annotations are added to help users better understand the algorithm.

Create a conda environment:

conda env create -f conda_env.yaml
conda activate act

python utils/collect_data.py

python train.py --dataset_dir <YOUR_DATASET_ROOT>

python eval.py --checkpoint <YOUR_CHECKPOINT_PATH>