Merge 'initializers' into 'TorchScriptGraph'

[BowenBao]

Reduce friction.

[codecov]

Codecov Report

Merging #480 (dcca04f) into main (5ba4fc7) will increase coverage by 0.00%.
The diff coverage is 90.90%.

@@           Coverage Diff           @@
##             main     #480   +/-   ##
=======================================
  Coverage   72.45%   72.46%           
=======================================
  Files         109      109           
  Lines       10602    10607    +5     
  Branches     1093     1093           
=======================================
+ Hits         7682     7686    +4     
- Misses       2613     2614    +1     
  Partials      307      307           

Impacted Files	Coverage Δ
...xscript/function_libs/torch_aten/graph_building.py	69.16% <87.50%> (+0.49%)	⬆️
...pt/function_libs/torch_aten/graph_building_test.py	88.67% <100.00%> (-0.42%)	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[titaiwangms]

I think torch-nightly is broken by #460

[BowenBao]

@titaiwangms thx, hard to figure out what are known failures...

Does the CI here depend on code from torch exporter? This api change requires modification from both sides, I was planning to merge this first hoping CI could be all green.

[titaiwangms]

@titaiwangms thx, hard to figure out what are known failures...

Does the CI here depend on code from torch exporter? This api change requires modification from both sides, I was planning to merge this first hoping CI could be all green.

I don't think onnx-script relies on torch exporter, but what part in this PR needs torch exporter modification? As far as I know, the experimental branch is broken due to lack of maintain on ort-experimental branch, and torch-nightly fails because of squeeze op. There are other two erroring with ORT that I don't quite understand, but this PR seems to be unrelated...

[BowenBao]

The api changed, initializers no longer an argument to TorchScripGraph.to_model_proto, and pt exporter need to adjust for it, which is in pytorch/pytorch#95676. It does seem that pt exporter is not involved in any of the CI tests here.

[justinchuby]

Fixes #389

[justinchuby]

Would a brief docstring for this property help?

[justinchuby]

The model opset can actually be different from the local functions' opsets I think. Do you know if that's true?

[BowenBao]

Yea it could be different under constraint.
https://github.com/onnx/onnx/blob/4b2d50334914621835cc1e8dadd4fe82b6b9876c/onnx/onnx.in.proto#L824-L828

  // The operator sets imported by FunctionProto should be compatible with the ones
  // imported by ModelProto. Example, if same operator set say 'A' is imported by FunctionProto
  // and ModelProto then versions for the operator set may be different but,
  // the operator schema returned for op_type, domain, version combination
  // for both the versions should be same.