[bug-fix] enforce all converted quantized ONNX initializers are same dtype #1181
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
KSGulin
approved these changes
Nov 23, 2022
There was a problem hiding this comment.
Tested the scenarios:
- Torch==1.12.1 +
quantize_conv_activations: Falsein quantization modifier - Torch==1.12.1 + NO
quantize_conv_activations: Falsein quantization modifier - Torch==1.9.1 +
quantize_conv_activations: Falsein quantization modifier
All passed. Great fix! 🥳
bfineran
added a commit
that referenced
this pull request
Nov 23, 2022
bfineran
added a commit
that referenced
this pull request
Nov 23, 2022
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Outdated suggestions cannot be applied.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.
torch 1.12 exports some Q/DQ inputs as initializers now instead of constants - this causes a pass in the quantized graph conversion step to miss a conversion from INT8 to UINT8. this PR adds an explicit pass to convert any missed INT8 to UINT8 initializers. without this fix, models may fail on ORT and deepsparse compile due to a mismatch between INT8 and UINT8 inputs
test_plan:
@KSGulin has reproduced the error to verify fix and commandeer
need to verify compilation + accuracy