PPO tuning: LR anneal, value clipping, per-minibatch adv norm#128
Open
dnddnjs wants to merge 1 commit into
Open
PPO tuning: LR anneal, value clipping, per-minibatch adv norm#128dnddnjs wants to merge 1 commit into
dnddnjs wants to merge 1 commit into
Conversation
…ames Three of CleanRL's 'PPO 37 details' that were missing — flagged when the 5M and 10M Breakout runs both plateaued at per-game ~75 with entropy stuck around 0.8 (policy wasn't sharpening, clip rarely activating): - Linear LR anneal from 2.5e-4 -> 0 across the run; lets late updates fine-tune instead of bouncing. - Value-function loss clipping around the old prediction (CLIP_COEF), matching the policy clipping range; stabilizes value targets. - Advantage normalization moved inside the minibatch loop instead of once per batch. Also bumps TOTAL_FRAMES 5M -> 10M to match the CleanRL Atari budget so runs are directly comparable to their published curves. lr now logged to wandb so the anneal is visible.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Three CleanRL 'PPO 37 details' that were missing, plus a frame budget bump:
lrto W&B so the anneal is visible.Why
5M and 10M Breakout runs both plateaued at per-game ~75, with entropy stuck around 0.8 and the PPO clip rarely activating — classic 'missing details' symptoms. These three changes are the standard fixes; the audit against CleanRL
ppo_atari.pyflagged exactly these.Test plan
TOTAL_FRAMES=5120) succeeds