Alexander Suslov

@dreaquil The following PR #1220 should fix the issue. Stay tuned!

jenkins retest this please

jenkins retest this please

Hi, @lpuglia. Thanks for your feedback and proposal! - The `multistep` scheduler ignores `sparsity_target` and `sparsity_target_epoch` parameters and calculates these parameters from `multistep_steps` and `multistep_sparsity_levels` parameters. Based on your feedback...

Hello @moonlightian, I think you can understand that the knowledge distillation works by the compression loss. It also includes the knowledge distillation loss. BTW, The `resnet34_pruning_geometric_median_kd.json` has the `"pruning_steps": 20`...

> Results for the evaluation:  > > Result for the openvino IR model evaluation(Mask model AC not supported): > Model AC Measured Reference FP32(OV) Diff FP32 within Threshold >...

> > > Results for the evaluation:  > > > Result for the openvino IR model evaluation(Mask model AC not supported): > > > Model AC Measured Reference FP32(OV)...

> Please refer to this build NNCF/job/nightly/job/e2e_onnx_model_zoo/22/ about this change. Thanks for the link. Could you make the report in the same style as all e2e builds?

> Which one do you mean the table schema, cell color formatting, or text font? I mean something like that https://github.com/openvinotoolkit/nncf/blob/cf8784f3811c40c3aa99ba63593638589472542b/tests/tensorflow/test_sota_checkpoints.py#L200

Hello @luoyger! The provided code looks pretty good. Could you share with me model and pctr_int8.json file to reproduce on my side?