fix V-trace importance weighting in CPU and scalar#598
Open
valtterivalo wants to merge 1 commit into
Open
Conversation
…dvantage The scalar CUDA kernel and the CPU advantage applied the clipped importance weight to the reward term only (rho*r + gamma*V' - V). The vectorized CUDA kernel already applies it to the whole TD error (rho*(r + gamma*V' - V)), the canonical V-trace form. The three paths only agree on-policy (rho == 1), so off-policy advantage estimates diverged on the scalar-routed and CPU paths. Align scalar and CPU with the vectorized kernel.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
missing parentheses here in CPU and scalar kernels. vec is good alrdy.