gryf/coach
1
0
Fork 0
mirror of https://github.com/gryf/coach.git synced 2025-12-18 11:40:18 +01:00
Code Issues Projects Releases Wiki Activity

Added average total reward to logging after evaluation phase completes. (#93)

Browse Source
This commit is contained in:
Thom Lane
2018-11-16 08:22:00 -08:00
committed by Scott Leishman
parent 81bac050d7
commit a0f25034c3
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
rl_coach/agents/agent.py
View File

@@ -367,9 +367,9 @@ class Agent(AgentInterface):
# we write to the next episode, because it could be that the current episode was already written # we write to the next episode, because it could be that the current episode was already written
# to disk and then we won't write it again # to disk and then we won't write it again
self.agent_logger.set_current_time(self.current_episode + 1) self.agent_logger.set_current_time(self.current_episode + 1)
evaluation_reward = self.accumulated_rewards_across_evaluation_episodes / self.num_evaluation_episodes_completed
self.agent_logger.create_signal_value( self.agent_logger.create_signal_value(
'Evaluation Reward', 'Evaluation Reward', evaluation_reward)
self.accumulated_rewards_across_evaluation_episodes / self.num_evaluation_episodes_completed)
self.agent_logger.create_signal_value( self.agent_logger.create_signal_value(
'Shaped Evaluation Reward', 'Shaped Evaluation Reward',
self.accumulated_shaped_rewards_across_evaluation_episodes / self.num_evaluation_episodes_completed) self.accumulated_shaped_rewards_across_evaluation_episodes / self.num_evaluation_episodes_completed)
@@ -379,8 +379,8 @@ class Agent(AgentInterface):
success_rate success_rate
) )
if self.ap.is_a_highest_level_agent or self.ap.task_parameters.verbosity == "high": if self.ap.is_a_highest_level_agent or self.ap.task_parameters.verbosity == "high":
screen.log_title("{}: Finished evaluation phase. Success rate = {}" screen.log_title("{}: Finished evaluation phase. Success rate = {}, Avg Total Reward = {}"
.format(self.name, np.round(success_rate, 2))) .format(self.name, np.round(success_rate, 2), np.round(evaluation_reward, 2)))
def call_memory(self, func, args=()): def call_memory(self, func, args=()):
""" """