[ONNX OP] DynamicQuantizeMatMul #28158
Open
+372
−0
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
…ted nuances in op implementation
|
@gkrivor
PR is able to pass the test case on CPU / GPU both, but with slightly high tolerance value of 0.0055f. I believe this is due to slight nuance difference in ONNX RT implementation vs how I have implemented in OpenVINO. |
|
build_jenkins |
…-tensor quantization
|
Added new test case and model to check the bias functionality + per-tensor quantization. Both test cases passing with 0.0055f tolerance value.
Test case 1 model:
Test case 2 model:
|
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Outdated suggestions cannot be applied.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.
Pseudo-Implementation of DQMM ONNX Contrib Op (Mimic Functionality) for OpenVINO ONNX Frontend
This PR does not implement DQMM exactly as ONNX RT - e.g. MatMulInteger is not used, instead ov::MatMul is used on dequantized matrix B
Please do not merge yet