hiro/ray
1
0
Fork 0
mirror of https://github.com/vale981/ray synced 2025-03-06 18:41:40 -05:00
Code Issues Projects Releases Packages Wiki Activity Actions

[rllib] Use NoFilter instead of MeanStdFilter for PPO. (#1082)

Browse source 
* Make NoFilter the default observation filter for PPO.

* Make reward filter NoFilter for PPO.
This commit is contained in:
Robert Nishihara 2017-10-04 21:31:17 -07:00 • committed by Philipp Moritz
parent a72084c568
commit 971becc905
2 changed files with 2 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
python/ray/rllib/ppo/ppo.py
View file

@ -56,7 +56,7 @@ DEFAULT_CONFIG = {
# Config params to pass to the model # Config params to pass to the model
"model": {"free_log_std": False}, "model": {"free_log_std": False},
# Which observation filter to apply to the observation # Which observation filter to apply to the observation
"observation_filter": "MeanStdFilter", "observation_filter": "NoFilter",
# If >1, adds frameskip # If >1, adds frameskip
"extra_frameskip": 1, "extra_frameskip": 1,
# Number of timesteps collected in each outer loop # Number of timesteps collected in each outer loop

2
python/ray/rllib/ppo/runner.py
View file

@ -148,7 +148,7 @@ class Runner(object):
else: else:
raise Exception("Unknown observation_filter: " + raise Exception("Unknown observation_filter: " +
str(config["observation_filter"])) str(config["observation_filter"]))
self.reward_filter = MeanStdFilter((), clip=5.0) self.reward_filter = NoFilter()
self.sess.run(tf.global_variables_initializer()) self.sess.run(tf.global_variables_initializer())
def load_data(self, trajectories, full_trace): def load_data(self, trajectories, full_trace):