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

Multiple improvements and bug fixes (#66)

Browse Source 
* Multiple improvements and bug fixes:

    * Using lazy stacking to save on memory when using a replay buffer
    * Remove step counting for evaluation episodes
    * Reset game between heatup and training
    * Major bug fixes in NEC (is reproducing the paper results for pong now)
    * Image input rescaling to 0-1 is now optional
    * Change the terminal title to be the experiment name
    * Observation cropping for atari is now optional
    * Added random number of noop actions for gym to match the dqn paper
    * Fixed a bug where the evaluation episodes won't start with the max possible ale lives
    * Added a script for plotting the results of an experiment over all the atari games
This commit is contained in:
Itai Caspi
2018-02-26 12:29:07 +02:00
committed by GitHub
parent 4fe9cba445
commit a7206ed702
20 changed files with 465 additions and 158 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
memories/episodic_experience_replay.py
View File

@@ -16,6 +16,7 @@
from memories.memory import *
import threading
from typing import Union
class EpisodicExperienceReplay(Memory):
@@ -103,7 +104,8 @@ class EpisodicExperienceReplay(Memory):
if transition.game_over:
self._num_transitions_in_complete_episodes += last_episode.length()
self._length += 1
self.buffer[-1].update_returns(self.discount, is_bootstrapped=self.return_is_bootstrapped,
self.buffer[-1].update_returns(self.discount,
is_bootstrapped=self.tp.agent.bootstrap_total_return_from_old_policy,
n_step_return=self.tp.agent.n_step)
self.buffer[-1].update_measurements_targets(self.tp.agent.num_predicted_steps_ahead)
# self.buffer[-1].update_actions_probabilities() # used for off-policy policy optimization
@@ -146,6 +148,17 @@ class EpisodicExperienceReplay(Memory):
def get(self, index):
return self.get_episode(index)
def get_last_complete_episode(self) -> Union[None, Episode]:
"""
Returns the last complete episode in the memory or None if there are no complete episodes
:return: None or the last complete episode
"""
last_complete_episode_index = self.num_complete_episodes()-1
if last_complete_episode_index >= 0:
return self.get(last_complete_episode_index)
else:
return None
def update_last_transition_info(self, info):
episode = self.buffer[-1]
if episode.length() == 0: