Steam Community :: Vekklern

Vekklern

pls rember that wen u feel scare or frigten
never forget ttimes wen u feeled happy
wen day is dark alway rember happy day

View more info

Currently Offline

Badges 27

Tokki
Offline

禮安
Online

Yogi (ヨギ・ベア)
Online

Kopciuch
Offline

Trelemorele
Offline

NefariousGoblin
In-Game
S.T.A.L.K.E.R. 2: Heart of Chornobyl

Game Collector

Featured Games

Favorite Game

STAR WARS™ Knights of the Old Republic™

Hours played

Review 1

Background for PPO

PPO is motivated by the same question as TRPO: how can we take the biggest possible improvement step on a policy using the data we currently have, without stepping so far that we accidentally cause performance collapse? Where TRPO tries to solve this problem with a complex second-order method, PPO is a family of first-order methods that use a few other tricks to keep new policies close to old. PPO methods are significantly simpler to implement, and empirically seem to perform at least as well as TRPO.

There are two primary variants of PPO: PPO-Penalty and PPO-Clip.

PPO-Penalty approximately solves a KL-constrained update like TRPO, but penalizes the KL-divergence in the objective function instead of making it a hard constraint, and automatically adjusts the penalty coefficient over the course of training so that it’s scaled appropriately.

PPO-Clip doesn’t have a KL-divergence term in the objective and doesn’t have a constraint at all. Instead relies on specialized clipping in the objective function to remove incentives for the new policy to get far from the old policy.

Recent Activity

10.5 hours past 2 weeks