Add AI Evaluation tests for eShopSupport.

[peterwald]

Add EvaluationTests project that demonstrates how to use Microsoft.Extensions.AI.Evaluation to run evaluations on LLM prompts and responses.

[shyamnamboodiripad]

@peterwald Should we add a README.md file under the test/EvaluationTests folder that contains instructions for running the tests / viewing the report?

[shyamnamboodiripad]

@luisquintanilla @peterwald should we also remove the src\Evaluator project now that we have EvaluationTests? This could be something to look into later on (post merge) as this could help to make the guidance on how to run evaluations less ambiguous overall.

I am guessing we don't want to do this yet since not all tests from the src\Evaluator have been ported over yet?

[luisquintanilla]

@luisquintanilla @peterwald should we also remove the src\Evaluator project now that we have EvaluationTests? This could be something to look into later on (post merge) as this could help to make the guidance on how to run evaluations less ambiguous overall.

I am guessing we don't want to do this yet since not all tests from the src\Evaluator have been ported over yet?

Since we're recommending using the evaluations in the test directory, I think it's fine to remove it.

[peterwald]

@luisquintanilla @peterwald should we also remove the src\Evaluator project now that we have EvaluationTests? This could be something to look into later on (post merge) as this could help to make the guidance on how to run evaluations less ambiguous overall.
I am guessing we don't want to do this yet since not all tests from the src\Evaluator have been ported over yet?

Since we're recommending using the evaluations in the test directory, I think it's fine to remove it.

There is more in the Evaluator project than what is done by the tests. I'll leave it here for now to not eliminate those options.

[shyamnamboodiripad]

@luisquintanilla @peterwald should we also remove the src\Evaluator project now that we have EvaluationTests? This could be something to look into later on (post merge) as this could help to make the guidance on how to run evaluations less ambiguous overall.
I am guessing we don't want to do this yet since not all tests from the src\Evaluator have been ported over yet?

Since we're recommending using the evaluations in the test directory, I think it's fine to remove it.

There is more in the Evaluator project than what is done by the tests. I'll leave it here for now to not eliminate those options.

Sounds good - should we log an issue to port over the remaining functionality and remove this eventually? @luisquintanilla

[luisquintanilla]

Sounds good - should we log an issue to port over the remaining functionality and remove this eventually? @luisquintanilla

Yeah. That would be great.

Tracking here #50. Feel free to update / add details as needed.

[luisquintanilla]

LGTM. Thanks @peterwald @shyamnamboodiripad

One last question before merging. Is this using the latest version of the libraries? If not, can you bump the version and then I'll merge.

Thanks