gaspersic/baselines
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Adds description of param_noise parameter in deepq.learn method (#675)

Browse Source
This commit is contained in:
Juliano Laganá
2018-10-24 19:00:31 +02:00
committed by pzhokhov
parent 84ea7aa1fd
commit c3bd8cea66
1 changed files with 2 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
baselines/deepq/deepq.py
View File

@@ -169,6 +169,8 @@ def learn(env,
to 1.0. If set to None equals to total_timesteps. to 1.0. If set to None equals to total_timesteps.
prioritized_replay_eps: float prioritized_replay_eps: float
epsilon to add to the TD errors when updating priorities. epsilon to add to the TD errors when updating priorities.
param_noise: bool
whether or not to use parameter space noise (https://arxiv.org/abs/1706.01905)
callback: (locals, globals) -> None callback: (locals, globals) -> None
function called at every steps with state of the algorithm. function called at every steps with state of the algorithm.
If callback returns true training stops. If callback returns true training stops.