Add DataViewSchema overloads to ConvertToOnnx

[FranklinWhale]

Fixes #6448

[codecov]

Codecov Report

Merging #6449 (0180ff6) into main (87337c0) will decrease coverage by 0.05%.
The diff coverage is 0.00%.

❗ Current head 0180ff6 differs from pull request most recent head aed491f. Consider uploading reports for the commit aed491f to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6449      +/-   ##
==========================================
- Coverage   68.55%   68.49%   -0.06%     
==========================================
  Files        1170     1170              
  Lines      246977   246980       +3     
  Branches    25792    25792              
==========================================
- Hits       169303   169164     -139     
- Misses      70951    71075     +124     
- Partials     6723     6741      +18     

Flag	Coverage Δ
Debug	68.49% <0.00%> (-0.06%)	⬇️
production	62.87% <0.00%> (-0.08%)	⬇️
test	88.96% <ø> (ø)

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

Impacted Files	Coverage Δ
...Microsoft.ML.OnnxConverter/OnnxExportExtensions.cs	91.42% <0.00%> (-8.58%)	⬇️
...osoft.ML.KMeansClustering/KMeansPlusPlusTrainer.cs	83.92% <0.00%> (-7.16%)	⬇️
src/Microsoft.ML.FastTree/Training/StepSearch.cs	57.42% <0.00%> (-4.96%)	⬇️
src/Microsoft.ML.Data/Training/TrainerUtils.cs	66.26% <0.00%> (-3.82%)	⬇️
...crosoft.ML.StandardTrainers/Standard/SdcaBinary.cs	85.40% <0.00%> (-3.42%)	⬇️
src/Microsoft.ML.Sweeper/AsyncSweeper.cs	71.42% <0.00%> (-1.37%)	⬇️
...oft.ML.StandardTrainers/Standard/SdcaMulticlass.cs	91.46% <0.00%> (-1.03%)	⬇️
src/Microsoft.ML.Data/Utils/LossFunctions.cs	66.83% <0.00%> (-0.52%)	⬇️
...soft.ML.Data/DataLoadSave/Text/TextLoaderCursor.cs	89.61% <0.00%> (-0.16%)	⬇️
...ML.Transforms/Text/StopWordsRemovingTransformer.cs	86.38% <0.00%> (-0.15%)	⬇️
... and 1 more

[FranklinWhale]

No idea why UseKeyDataViewTypeAsUInt32InOnnxInput keeps failing... The exception occurs during cursor disposal.

[michaelgsharp]

I added a comment in the issue itself, but I'll add it here as well.

So the call to Transform itself is a lazy call. Unless you get a cursor and start itterating over the data itself this call should basically be a NO-OP (not quite true as it does some validations and still sets up the pipeline, but no actual "processing" should take place).

Are you seeing something that is showing its actually doing data processing there?

[michaelgsharp]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).

[FranklinWhale]

Actually the original issue is about more the need to convert an ITransformer to ONNX with DataViewSchema only, because the data that trained the model is not always available (e.g. the model is imported).

However, the failed test case UseKeyDataViewTypeAsUInt32InOnnxInput seems to suggest that some data processing was involved during the conversion.