gryf/coach
1
0
Fork 0
mirror of https://github.com/gryf/coach.git synced 2025-12-17 19:20:19 +01:00
Code Issues Projects Releases Wiki Activity

Itaicaspi/episode reset refactoring (#105)

Browse Source 
* reordering of the episode reset operation and allowing to store episodes only when they are terminated

* reordering of the episode reset operation and allowing to store episodes only when they are terminated

* revert tensorflow-gpu to 1.9.0 + bug fix in should_train()

* tests readme file and refactoring of policy optimization agent train function

* Update README.md

* Update README.md

* additional policy optimization train function simplifications

* Updated the traces after the reordering of the environment reset

* docker and jenkins files

* updated the traces to the ones from within the docker container

* updated traces and added control suite to the docker

* updated jenkins file with the intel proxy + updated doom basic a3c test params

* updated line breaks in jenkins file

* added a missing line break in jenkins file

* refining trace tests ignored presets + adding a configurable beta entropy value

* switch the order of trace and golden tests in jenkins + fix golden tests processes not killed issue

* updated benchmarks for dueling ddqn breakout and pong

* allowing dynamic updates to the loss weights + bug fix in episode.update_returns

* remove docker and jenkins file
This commit is contained in:
Itai Caspi
2018-09-04 15:07:54 +03:00
committed by GitHub
parent 7086492127
commit 72a1d9d426
92 changed files with 9803 additions and 9740 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
rl_coach/level_manager.py
View File

@@ -70,7 +70,7 @@ class LevelManager(EnvironmentInterface):
self.should_reset_agent_state_after_time_limit_passes = should_reset_agent_state_after_time_limit_passes
self.full_name_id = self.name = name
self._phase = RunPhase.HEATUP
self.level_was_reset = True
self.reset_required = False
# set self as the parent for all the composite agents
for agent in self.agents.values():
@@ -104,7 +104,7 @@ class LevelManager(EnvironmentInterface):
:return: the environment response as returned in get_last_env_response
"""
[agent.reset_internal_state() for agent in self.agents.values()]
self.level_was_reset = True
self.reset_required = False
if self.real_environment.current_episode_steps_counter == 0:
self.last_env_response = self.real_environment.last_env_response
return self.last_env_response
@@ -203,6 +203,9 @@ class LevelManager(EnvironmentInterface):
for agent_name, agent in self.agents.items():
agent.set_incoming_directive(action)
if self.reset_required:
self.reset_internal_state()
# get last response or initial response from the environment
env_response = copy.copy(self.environment.last_env_response)
@@ -238,7 +241,7 @@ class LevelManager(EnvironmentInterface):
# this is the agent's only opportunity to observe this transition - he will not get another one
acting_agent.observe(env_response) # TODO: acting agent? maybe all of the agents in the layer?
self.handle_episode_ended()
self.reset_internal_state()
self.reset_required = True
return env_response_for_upper_level