Upgrade onnxsim to v0.4.0

[daquexian]

Motivation

I'm the author of onnx-simplifier. Recently I published onnxsim v0.4.0. In the latest version, dynamic input shapes are natively supported with no need for extra flags like dynamic_input_shapes. custom_lib and input_data is also not needed.

[CLAassistant]

All committers have signed the CLA.

[codecov[bot]]

Codecov Report

Merging #915 (3f71c11) into master (d2e5054) will decrease coverage by 0.64%. The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #915      +/-   ##
==========================================
- Coverage   85.46%   84.82%   -0.65%     
==========================================
  Files         132      132              
  Lines        8750     8749       -1     
  Branches     1513     1513              
==========================================
- Hits         7478     7421      -57     
- Misses       1050     1090      +40     
- Partials      222      238      +16     

Flag	Coverage Δ
unittests	84.82% <ø> (-0.57%)	:arrow_down:

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
mmcls/utils/setup_env.py	72.72% <0.00%> (-27.28%)	:arrow_down:
mmcls/models/utils/helpers.py	79.16% <0.00%> (-20.84%)	:arrow_down:
mmcls/models/backbones/convnext.py	80.00% <0.00%> (-14.79%)	:arrow_down:
mmcls/datasets/custom.py	88.52% <0.00%> (-11.48%)	:arrow_down:
mmcls/models/backbones/convmixer.py	88.52% <0.00%> (-9.84%)	:arrow_down:
mmcls/datasets/builder.py	78.37% <0.00%> (-9.46%)	:arrow_down:
mmcls/models/backbones/cspnet.py	88.23% <0.00%> (-4.58%)	:arrow_down:
mmcls/core/visualization/image.py	86.20% <0.00%> (-0.98%)	:arrow_down:
mmcls/datasets/dataset_wrappers.py	71.83% <0.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update d2e5054...3f71c11. Read the comment docs.

[mzr1996]

Thanks for your contribution. In fact, the deploy functionalities in mmclassification are deprecating, we move all these functionalities to mmdeploy. Have you informed them?