Wonju Lee
Wonju Lee
@ljaljushkin, could you review this?
test this please
You can download the models from https://storage.openvinotoolkit.org/repositories/nncf/models/develop/torch/. Thank you.
Instead of directly inferencing from .pth file, we provide some notebooks for inferencing with OpenVINO IR (convert pth -> onnx -> OpenVINO IR) to compare FP32 and INT8. Please refer...
> @vshampor @andrey-churkin While running the TF2.8 weekly training, mobilenet_v3 sparsity quantization test shows 67.61% and 67.62% accuracy which is better than expected result 67.55%, so that we can pass...
@sooahleex, could you take a look?
Hi @sh0622-kim, thank you for your interest in Datumaro. Since the new feature list is frozen for next Datumaro (v2.0.0), we are going to support this format in the 2.1.0...
Hi @sh0622-kim, DOTA format is now available in the develop branch! Please refer to https://github.com/openvinotoolkit/datumaro/pull/1475 and https://openvinotoolkit.github.io/datumaro/latest/docs/data-formats/formats/dota.html
Thank you for your interests @DP1701. From your report, we felt the need of data preparation for K-fold cross validation. We are going to get back with this feature in...
Hi @hugo082, sorry for the late response, sure, we are always open to add more features. Please feel free to create the PR. This will be helpful to be consistent...