Improve normalized advantage calculation #1642
Merged
Conversation
This file contains 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
1. Use LazyBatchNorm for advantage normalization.The motivation is that for DDP training, the normalization statistics will combine the statistics from all GPUs. 2. Calculate normalized advantage in PPOAlgorithm.preprocess so that the normalization is based on a much larger batch.
Not sure why computing normalized advantage in preprocess is based on a larger batch? Is it due to mini_batch_length smaller than full batch length? |
In preprocess, all the samples from the rollout are processed. In loss, only mini_batch_size samples are processed. |
hnyu
approved these changes
May 15, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Use LazyBatchNorm for advantage normalization.The motivation is that for DDP training, the normalization statistics will combine the statistics from all GPUs.
Calculate normalized advantage in PPOAlgorithm.preprocess so that the normalization is based on a much larger batch.