Implement GatherND operator
#155
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
Add a variant of the `inner_iter` iterators for the case where the number of inner dimensions is determined at runtime.
1 task
By ensuring the output and `indices` are contiguous, we can reduce dynamic-rank `TensorView` creation in the inner loop and operate directly on slices. This reduces time in GatherND ops in the Piper TTS model by ~50%, but this is only doing fairly small gather ops.
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Outdated suggestions cannot be applied.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.
Add an unoptimized
GatherNDimplementation, along with some newTensorBasemethods to support the implementation.This is one of three changes needed to run Piper TTS models, the others being support for 1D ConvTranspose with padding, and bug fixes for
ScatterND.TODO: