Arjun Suresh
Arjun Suresh
Sure @tjablin. Nothing urgent.
@pgmpablo157321 Sure. I'll close this PR and do another one with the updated changes.
This is the accuracy we got with the reference implementation (fp32) using onnxruntime backend. Not sure why but mAP is lower than 37.575% ( 37.572% ) ``` Average Precision (AP)...
My bad. Sorry I was careless and misread the decimal points. Below are the official figures as in r2.1 branch 1. Pytorch model: 0.3755 2. Onnx model: 0.3757 (0.37572 reproduced...
@tjablin @pgmpablo157321 Can you please approve this as this can fix a minor formatting issue on the results table?
Fixed from r2.1
I think "PASS_VARIATION_TAGS_TO_POST_DEPS" may not be a good solution as we might have multiple post_deps for a script. Instead we can have `add_to_deps` list inside each variation where the variation,...
yes Grigori. By adding in _cm.json we are documenting them also :)
Part of another PR. Hence closing
Hi @psyhtest , are you referring to [this script](https://github.com/mlcommons/inference/blob/master/vision/classification_and_detection/tools/accuracy-openimages.py)?