[mlir-tensorrt] Raise activations from their elementwise representation to tensorrt.activation Op #679
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
matthewfl
requested review from
christopherbate and
shelkesagar29
as code owners
matthewfl
force-pushed
the
mfl/raise-activations
branch
from
1d549c7 to
83014d4
Compare
mlir-tensorrt/tensorrt/lib/TensorRT/Transforms/RaiseActivations.pdll
Outdated
Show resolved
Hide resolved
matthewfl
force-pushed
the
mfl/raise-activations
branch
from
0685a68 to
221b51a
Compare
…on to tensorrt.activation Op adds new raising from `tensorrt.elemetnwise` Ops to `tensorrt.activation` for the GELU Tanh (when created by torch-mlir) and GELU Erf. This also includes raising from `min(a, max(x, y))` to the CLIP activation (which is used by clip and ReLU6 by torch-mlir). Merging these elementwise ops in to the `tensorrt.activation` type enables for TensorRT to fuse the activation into a proceeding linear layer's kernel (matrix multiply + elementwise sum). Signed-off-by: Matthew Francis-Landau <[email protected]> make raise activation use same getSplatConstantElementAttribute utils function Signed-off-by: Matthew Francis-Landau <[email protected]>
matthewfl
force-pushed
the
mfl/raise-activations
branch
from
221b51a to
ba10b6d
Compare
shelkesagar29
approved these changes
Aug 7, 2025
Contributor
Author
|
@shelkesagar29 can you merge the PR. (I don't have a merge button that I can click) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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Outdated suggestions cannot be applied.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.
This PR adds new raising from
tensorrt.elemetnwiseOps totensorrt.activationfor the GELU Tanh (when created by torch-mlir) and GELU Erf. This also includes raising frommin(a, max(x, y))to the CLIP activation (which is used by clip and ReLU6 by torch-mlir). Merging these elementwise ops in to thetensorrt.activationtype enables for TensorRT to fuse the activation into a proceeding linear layer's kernel (matrix multiply + elementwise sum).