gryf/coach
1
0
Fork 0
mirror of https://github.com/gryf/coach.git synced 2025-12-18 03:30:19 +01:00
Code Issues Projects Releases Wiki Activity

update CartPole_PPO not addressed during rebase (#41)

Browse Source
This commit is contained in:
Zach Dwiel
2018-10-24 19:58:25 -04:00
committed by Scott Leishman
parent f835ac902c
commit 2cc6abc3c4
1 changed files with 6 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
rl_coach/presets/CartPole_PPO.py
View File

@@ -1,9 +1,8 @@
from rl_coach.agents.clipped_ppo_agent import ClippedPPOAgentParameters
from rl_coach.architectures.tensorflow_components.architecture import Dense
from rl_coach.architectures.tensorflow_components.layers import Dense
from rl_coach.base_parameters import VisualizationParameters, PresetValidationParameters
from rl_coach.core_types import TrainingSteps, EnvironmentEpisodes, EnvironmentSteps, RunPhase
from rl_coach.environments.environment import MaxDumpMethod, SelectedPhaseOnlyDumpMethod, SingleLevelSelection
from rl_coach.environments.gym_environment import Mujoco, mujoco_v2, MujocoInputFilter
from rl_coach.environments.gym_environment import GymVectorEnvironment, mujoco_v2
from rl_coach.exploration_policies.additive_noise import AdditiveNoiseParameters
from rl_coach.exploration_policies.e_greedy import EGreedyParameters
from rl_coach.filters.observation.observation_normalization_filter import ObservationNormalizationFilter
@@ -29,8 +28,8 @@ agent_params = ClippedPPOAgentParameters()
agent_params.network_wrappers['main'].learning_rate = 0.0003
agent_params.network_wrappers['main'].input_embedders_parameters['observation'].activation_function = 'tanh'
agent_params.network_wrappers['main'].input_embedders_parameters['observation'].scheme = [Dense([64])]
agent_params.network_wrappers['main'].middleware_parameters.scheme = [Dense([64])]
agent_params.network_wrappers['main'].input_embedders_parameters['observation'].scheme = [Dense(64)]
agent_params.network_wrappers['main'].middleware_parameters.scheme = [Dense(64)]
agent_params.network_wrappers['main'].middleware_parameters.activation_function = 'tanh'
agent_params.network_wrappers['main'].batch_size = 64
agent_params.network_wrappers['main'].optimizer_epsilon = 1e-5
@@ -45,22 +44,15 @@ agent_params.algorithm.optimization_epochs = 10
agent_params.algorithm.estimate_state_value_using_gae = True
agent_params.algorithm.num_steps_between_copying_online_weights_to_target = EnvironmentSteps(2048)
# agent_params.input_filter = MujocoInputFilter()
agent_params.exploration = EGreedyParameters()
agent_params.exploration.epsilon_schedule = LinearSchedule(1.0, 0.01, 10000)
# agent_params.pre_network_filter = MujocoInputFilter()
agent_params.pre_network_filter.add_observation_filter('observation', 'normalize_observation',
ObservationNormalizationFilter(name='normalize_observation'))
###############
# Environment #
###############
env_params = Mujoco()
env_params.level = 'CartPole-v0'
visualization_params = VisualizationParameters()
visualization_params.video_dump_methods = [SelectedPhaseOnlyDumpMethod(RunPhase.TEST), MaxDumpMethod()]
visualization_params.dump_mp4 = False
env_params = GymVectorEnvironment(level='CartPole-v0')
########
# Test #
@@ -71,5 +63,5 @@ preset_validation_params.min_reward_threshold = 150
preset_validation_params.max_episodes_to_achieve_reward = 250
graph_manager = BasicRLGraphManager(agent_params=agent_params, env_params=env_params,
schedule_params=schedule_params, visualization_params=visualization_params,
schedule_params=schedule_params, vis_params=VisualizationParameters(),
preset_validation_params=preset_validation_params)