diff --git a/.github/workflows/run_algs.yaml b/.github/workflows/run_algs.yaml index b586bd7..0fa05c3 100644 --- a/.github/workflows/run_algs.yaml +++ b/.github/workflows/run_algs.yaml @@ -13,7 +13,7 @@ jobs: test-algorithms: name: "Python ${{ matrix.python-version }} on ${{ matrix.os }}" runs-on: "${{ matrix.os }}" - timeout-minutes: 30 + timeout-minutes: 60 strategy: matrix: diff --git a/bash_scripts/run-algorithms.sh b/bash_scripts/run-algorithms.sh index 3aabdb4..0b3120c 100644 --- a/bash_scripts/run-algorithms.sh +++ b/bash_scripts/run-algorithms.sh @@ -4,6 +4,7 @@ echo "Running All Algorithms..." +# Test a subset of algorithms python stoix/systems/ppo/anakin/ff_ppo.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 python stoix/systems/ppo/anakin/ff_ppo_continuous.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 python stoix/systems/q_learning/ff_dqn.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 @@ -17,3 +18,8 @@ python stoix/systems/ddpg/ff_td3.py arch.total_timesteps=300 arch.total_num_envs python stoix/systems/vpg/ff_reinforce.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 python stoix/systems/awr/ff_awr.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 python stoix/systems/mpo/ff_mpo.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 + +# Test a subset of networks +python stoix/systems/ppo/anakin/ff_ppo.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 env=gymnax/breakout env.wrapper=null network=cnn network.actor_network.pre_torso.channel_first=False network.critic_network.pre_torso.channel_first=False +python stoix/systems/ppo/anakin/ff_ppo.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 env=gymnax/breakout env.wrapper=null network=visual_resnet network.actor_network.pre_torso.channel_first=False network.critic_network.pre_torso.channel_first=False +python stoix/systems/ppo/anakin/ff_ppo.py arch.total_timesteps=300 arch.total_num_envs=8 arch.num_evaluation=1 system.rollout_length=8 network=mlp_resnet diff --git a/stoix/utils/make_env.py b/stoix/utils/make_env.py index 65821a8..37812f7 100644 --- a/stoix/utils/make_env.py +++ b/stoix/utils/make_env.py @@ -286,7 +286,7 @@ def apply_optional_wrappers( A tuple of the environments. """ envs = list(envs) - if "wrapper" in config.env: + if "wrapper" in config.env and config.env.wrapper is not None: for i in range(len(envs)): envs[i] = hydra.utils.instantiate(config.env.wrapper, env=envs[i])