Skip to content

TRAIN_EVAL.md

Latest commit

 

History

History
649 lines (615 loc) · 20.4 KB

TRAIN_EVAL.md

File metadata and controls

649 lines (615 loc) · 20.4 KB
Raw

Train and Evaluation

This file provides the CLIs we used to produce the data that present in the paper

Tip

To run the Meta-World with KUKA manipulation, add following command (Only valid for MT10_KUKA)

env.use_kuka_env=True

To run the LEGION under multi-task setting

python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=0 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=1 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=2 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=3 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=4 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=5 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=6 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=7 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=8 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000 
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=False \
env.use_onehot=False \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.encoder.vae.latent_dim=10 \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=6000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=50000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=30 \
agent.multitask.dpmm_cfg.beta_kl_z=0.002 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.num_envs=10 \
experiment.training_mode=multitask \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
experiment.save_video=False \
setup.seed=9 \
setup.device=cuda:0 \
replay_buffer.type_to_select=buffer \
replay_buffer.buffer.batch_size=1280 \
replay_buffer.buffer.capacity=1000000

To run the LEGION under lifelong learning setting

python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=0 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=1 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=2 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=3 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=4 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=5 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=6 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=7 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=8 \
setup.device=cuda:0
python3 -u main.py \
setup=continuouslearning \
env=metaworld-mt10 \
env.use_onehot=False \
env.use_kuka_env=True \
agent=sac_dpmm \
agent.encoder.type_to_select=vae \
agent.encoder.vae.should_reconstruct=True \
agent.multitask.should_use_task_encoder=True \
agent.multitask.should_use_disentangled_alpha=True \
agent.multitask.encoder_input_setup=context_obs \
agent.multitask.dpmm_cfg.dpmm_update_start_step=10000 \
agent.multitask.dpmm_cfg.dpmm_update_freq=100000 \
agent.multitask.dpmm_cfg.kl_div_update_freq=50 \
agent.multitask.dpmm_cfg.sF=0.00001 \
agent.multitask.dpmm_cfg.beta_kl_z=0.001 \
experiment.training_mode=crl_queue \
experiment.should_reset_optimizer=True \
experiment.should_reset_critics=False \
experiment.should_reset_vae=False \
experiment.eval_freq=10000 \
experiment.num_eval_episodes=10 \
experiment.num_train_steps=1000000 \
agent.multitask.num_envs=10 \
experiment.save_video=False \
setup.seed=9 \
setup.device=cuda:0

Evaluation

All training and evaluation data are saved under logs/YYYY-MM-DD HH:MM:SS.

In each data log folder, you will see:

dpmm_model                      -- saved DPMM model and latent variable data
model                           -- saved LEGION ckpts
tb_logger_MM-DD_HH-MM           -- tensorboard logger
video                           -- saved video visualization 
config.json                     -- config setup file
eval.log                        -- log data for evaluation
log.jsonl                       -- log hyperparameters
train.log                       -- log data for training

Baselines

We use MTRL and Continual World for baselines comparison.

  • MTRL supports 8 different multi-task RL algorithms as described here.

  • Continual world supports various continual learning methods inpired from foundation machine learning domain, baseline algorithms described here.