Algorithmic Designs

Reproducing Results

Fixing the POLICY to be one of [bc_rnn_policy, bc_transformer_policy, bc_vilt_policy]. Select an algorithm from [base, er, agem, ewc, packnet] and run:

export CUDA_VISIBLE_DEVICES=GPU_ID && \
export MUJOCO_EGL_DEVICE_ID=GPU_ID && \
python lifelong/main.py seed=SEED \
                        benchmark_name=BENCHMARK 
                        policy=POLICY \
                        lifelong=ALGO

Assitive Note - Specific command generation

Algorithm: Select an algorithm base (Sequential) er (Experience Replay) ewc (Elastic Weight Consolidation) packnet (PackNet) single_task (Single Task) multitask (Multitask)
Benchmrak: Select a benchmark libero_spatial (LIBERO-Spatial) libero_object (LIBERO-Object) libero_goal (LIBERO-Goal) libero_90 (90 short-horizon tasks in LIBERO-100) libero_10 (LIBERO-Long, 10 long-horizon tasks in LIBERO-100)
GPU Id:
Random Seed

ResNet-RNN Policy (bc_rnn_policy)

ResNet-T Policy (bc_transformer_policy)

ViT-T Policy (bc_vilt_policy)