Improves API parameter naming consistency #96
Merged
Conversation
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
Renames q/k/v parameters to query/key/value in flash attention functions for better readability and standardization. Updates parameter documentation to reflect the new naming convention and fixes GQA condition description to use <= instead of <. Removes outdated footer reference to integration docs.
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR improves API consistency by renaming abbreviated parameter names in flash attention functions to their full descriptive names (q/k/v → query/key/value). This enhances code readability and follows standard naming conventions.
- Renames abbreviated parameter names to full descriptive names in function signatures
- Updates parameter documentation to reflect the new naming convention
- Fixes GQA condition description and removes outdated footer reference
Comment on lines
+86
to
+87
| - key: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H | ||
| - value: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H |
There was a problem hiding this comment.
The GQA condition description should clarify what 'H' refers to. It should specify 'H_kv <= num_heads' or reference the query tensor's head dimension to avoid ambiguity.
Suggested change
| - key: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H | |
| - value: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H | |
| - key: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H (number of query heads in the query tensor) | |
| - value: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H (number of query heads in the query tensor) |
Comment on lines
+86
to
+87
| - key: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H | ||
| - value: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H |
There was a problem hiding this comment.
The GQA condition description should clarify what 'H' refers to. It should specify 'H_kv <= num_heads' or reference the query tensor's head dimension to avoid ambiguity.
Suggested change
| - key: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H | |
| - value: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H | |
| - key: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H (number of query heads) | |
| - value: (B, K, H_kv, D). Same dtype/device as query; GQA when H_kv <= H (number of query heads) |
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.
Renames q/k/v parameters to query/key/value in flash attention functions for better readability and standardization.
Updates parameter documentation to reflect the new naming convention and fixes GQA condition description to use <= instead of <.
Removes outdated footer reference to integration docs.