Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[ONNX] Initial work to import pre-quantized ONNX Models #7802

Merged
merged 3 commits into from
Apr 7, 2021

Conversation

mbrookhart
Copy link
Contributor

This PR implements QuantlizeLinear, DequantizeLinear, and DynamicQuantizeLinear (which is deprecated but included for completeness).

Copy link
Contributor

@jwfromm jwfromm left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Awesome, excited to see this in action.

@@ -4165,15 +4165,8 @@ def verify_cumsum(indata, axis, exclusive=0, reverse=0, type="float32"):
"test_cumsum_2d_axis_0/",
"test_cumsum_2d_axis_1/",
"test_cumsum_2d_negative_axis/",
"test_dequantizelinear/",
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we run these tests on CI?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, we're running all of the pre-serialized node tests that ship with ONNX against CPU now, except what's skipped in this list. Working on reducing what we skip, and I'll start enabling GPU soon.

@masahi masahi merged commit 2d3f781 into apache:main Apr 7, 2021
@masahi
Copy link
Member

masahi commented Apr 7, 2021

thanks @mbrookhart @jwfromm

trevor-m pushed a commit to trevor-m/tvm that referenced this pull request May 6, 2021
* Add QuantizeLinear and DequantizeLinear

* DynamicDequantizeLinear
trevor-m pushed a commit to trevor-m/tvm that referenced this pull request May 6, 2021
* Add QuantizeLinear and DequantizeLinear

* DynamicDequantizeLinear
trevor-m pushed a commit to trevor-m/tvm that referenced this pull request May 6, 2021
* Add QuantizeLinear and DequantizeLinear

* DynamicDequantizeLinear
trevor-m pushed a commit to neo-ai/tvm that referenced this pull request May 11, 2021
* Add QuantizeLinear and DequantizeLinear

* DynamicDequantizeLinear
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants