[Evals API][6/n] meta-reference llm as judge, registration for ScoringFnDefs (#330)

* wip scoring refactor

* llm as judge, move folders

* test full generation + eval

* extract score regex to llm context

* remove prints, cleanup braintrust in this branch

* change json -> class

* remove initialize

* address nits

* check identifier prefix

* udpate MANIFEST

llama_stack/providers/tests/datasetio/test_datasetio.py

@@ -70,6 +70,7 @@ async def register_dataset(
     if for_generation:
         dataset_schema = {
             "expected_answer": StringType(),
+            "input_query": StringType(),
             "chat_completion_input": ChatCompletionInputType(),
         }
     else: