Minor correction in 'Add & Norm' logic in Block Class in gpt.py#22
Open
AbhishekAshokDubey wants to merge 1 commit intokarpathy:masterfrom
Open
Minor correction in 'Add & Norm' logic in Block Class in gpt.py#22AbhishekAshokDubey wants to merge 1 commit intokarpathy:masterfrom
AbhishekAshokDubey wants to merge 1 commit intokarpathy:masterfrom
Conversation
Updating the forward function in Transformer block. The code is simple to example the pull request, but still trying my best to explain below: As per paper: In 'Add & Norm' block of Transformer, Layer Norm is applied on top of input/ residual & output of Self-attention. While in the current code, first layer Norm is applied & then added back to the input/ residual.
AbhishekAshokDubey
changed the title
|
See 1:35:33 |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Updating the forward function in Transformer block.
The change is simple, but still trying my best to explain below:
As per original paper: In 'Add & Norm' block of Transformer, Layer Norm is applied on top of => input/ residual and output of Self-attention. While in the current code, layer Norm is applied first & then added back to the input/ residual.