Moved TensorFlow samples to its own directory in Samples project.

[zeahmed]

This PR addresses following issues

1. Moved TF sample to TensorFlow folder in samples project.
2. Renamed the files to meaningful name.
3. Added a text classification sample.
4. Updated TensorFlow.Redist nuget to latest in samples project.

@sfilipi, @shmoradims and @rogancarr, let me know if this move can be a problem for any documentation referencing it.

[codecov]

Codecov Report

Merging #2429 into master will decrease coverage by 0.03%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##           master    #2429      +/-   ##
==========================================
- Coverage   71.26%   71.22%   -0.04%     
==========================================
  Files         785      785              
  Lines      140946   141034      +88     
  Branches    16108    16112       +4     
==========================================
+ Hits       100440   100450      +10     
- Misses      36039    36116      +77     
- Partials     4467     4468       +1

Flag	Coverage Δ
#Debug	71.22% <0%> (-0.04%)	⬇️
#production	67.56% <0%> (-0.05%)	⬇️
#test	85.32% <ø> (ø)	⬆️

[Ivanidzo4ka]

[sfilipi]

Sorry I didn't notice it in the first pass, but I think you are trying to add yet another example of how to use the ScoreTensorFlowModel xtension, for text this time?

If that is the case, that is unecessary, because the API reference samples just need one example to illustrate how to use the API. Samples for the Various scenarios, and things you can do with the components go under the machienlearning-samples repo.

Now, if you can turn around the sample, in a way that illustrates one of the other ScoreTensorFlowModel extensions, that would be useful. #Resolved

[sfilipi]

🕐

[zeahmed]

I feel one sample is not enough to show the various different models that can be used with TensorFlowTransform in ML.NET. I am updating the sample to use the different extension of ScoreTensorFlowModel with inspecting the schema of inputs.

---

In reply to: 461202841 [](ancestors = 461202841)

[zeahmed]

public [](start = 8, length = 6)

This need to be public. Its helpful to get the schema without loading model again. #Resolved

[sfilipi]