Fix precommit check after moving to ruff (#927)

Lint check in main branch is failing. This fixes the lint check after we
moved to ruff in https://github.com/meta-llama/llama-stack/pull/921. We
need to move to a `ruff.toml` file as well as fixing and ignoring some
additional checks.

Signed-off-by: Yuan Tang <terrytangyuan@gmail.com>

llama_stack/providers/inline/scoring/basic/scoring.py

@@ -54,15 +54,11 @@ class BasicScoringImpl(
 
     async def list_scoring_functions(self) -> List[ScoringFn]:
         scoring_fn_defs_list = [
-            fn_def
-            for impl in self.scoring_fn_id_impls.values()
-            for fn_def in impl.get_supported_scoring_fn_defs()
+            fn_def for impl in self.scoring_fn_id_impls.values() for fn_def in impl.get_supported_scoring_fn_defs()
         ]
 
         for f in scoring_fn_defs_list:
-            assert f.identifier.startswith(
-                "basic"
-            ), "All basic scoring fn must have identifier prefixed with 'basic'! "
+            assert f.identifier.startswith("basic"), "All basic scoring fn must have identifier prefixed with 'basic'! "
 
         return scoring_fn_defs_list
 
@@ -76,9 +72,7 @@ class BasicScoringImpl(
         save_results_dataset: bool = False,
     ) -> ScoreBatchResponse:
         dataset_def = await self.datasets_api.get_dataset(dataset_id=dataset_id)
-        validate_dataset_schema(
-            dataset_def.dataset_schema, get_valid_schemas(Api.scoring.value)
-        )
+        validate_dataset_schema(dataset_def.dataset_schema, get_valid_schemas(Api.scoring.value))
 
         all_rows = await self.datasetio_api.get_rows_paginated(
             dataset_id=dataset_id,
@@ -108,12 +102,8 @@ class BasicScoringImpl(
                 raise ValueError(f"Scoring function {scoring_fn_id} is not supported.")
             scoring_fn = self.scoring_fn_id_impls[scoring_fn_id]
             scoring_fn_params = scoring_functions.get(scoring_fn_id, None)
-            score_results = await scoring_fn.score(
-                input_rows, scoring_fn_id, scoring_fn_params
-            )
-            agg_results = await scoring_fn.aggregate(
-                score_results, scoring_fn_id, scoring_fn_params
-            )
+            score_results = await scoring_fn.score(input_rows, scoring_fn_id, scoring_fn_params)
+            agg_results = await scoring_fn.aggregate(score_results, scoring_fn_id, scoring_fn_params)
             res[scoring_fn_id] = ScoringResult(
                 score_rows=score_results,
                 aggregated_results=agg_results,

llama_stack/providers/inline/scoring/basic/scoring_fn/equality_scoring_fn.py

@@ -32,9 +32,7 @@ class EqualityScoringFn(RegisteredBaseScoringFn):
         scoring_params: Optional[ScoringFnParams] = None,
     ) -> ScoringResultRow:
         assert "expected_answer" in input_row, "Expected answer not found in input row."
-        assert (
-            "generated_answer" in input_row
-        ), "Generated answer not found in input row."
+        assert "generated_answer" in input_row, "Generated answer not found in input row."
 
         expected_answer = input_row["expected_answer"]
         generated_answer = input_row["generated_answer"]

llama_stack/providers/inline/scoring/basic/scoring_fn/fn_defs/equality.py

@@ -18,7 +18,5 @@ equality = ScoringFn(
     provider_id="basic",
     provider_resource_id="equality",
     return_type=NumberType(),
-    params=BasicScoringFnParams(
-        aggregation_functions=[AggregationFunctionType.accuracy]
-    ),
+    params=BasicScoringFnParams(aggregation_functions=[AggregationFunctionType.accuracy]),
 )

llama_stack/providers/inline/scoring/basic/scoring_fn/fn_defs/regex_parser_multiple_choice_answer.py

@@ -55,9 +55,7 @@ MULTILINGUAL_ANSWER_REGEXES = [
     r"Àṣàyàn\s*:",
 ]
 
-MULTILINGUAL_ANSWER_PATTERN_TEMPLATE = (
-    r"(?i){}\s*([A-D]|[أ-د]|[অ]|[ব]|[ড]|[ঢ]|[Ａ]|[Ｂ]|[Ｃ]|[Ｄ])"
-)
+MULTILINGUAL_ANSWER_PATTERN_TEMPLATE = r"(?i){}\s*([A-D]|[أ-د]|[অ]|[ব]|[ড]|[ঢ]|[Ａ]|[Ｂ]|[Ｃ]|[Ｄ])"
 
 regex_parser_multiple_choice_answer = ScoringFn(
     identifier="basic::regex_parser_multiple_choice_answer",
@@ -66,10 +64,7 @@ regex_parser_multiple_choice_answer = ScoringFn(
     provider_id="basic",
     provider_resource_id="regex-parser-multiple-choice-answer",
     params=RegexParserScoringFnParams(
-        parsing_regexes=[
-            MULTILINGUAL_ANSWER_PATTERN_TEMPLATE.format(x)
-            for x in MULTILINGUAL_ANSWER_REGEXES
-        ],
+        parsing_regexes=[MULTILINGUAL_ANSWER_PATTERN_TEMPLATE.format(x) for x in MULTILINGUAL_ANSWER_REGEXES],
         aggregation_functions=[AggregationFunctionType.accuracy],
     ),
 )

llama_stack/providers/inline/scoring/basic/scoring_fn/fn_defs/subset_of.py

@@ -18,7 +18,5 @@ subset_of = ScoringFn(
     return_type=NumberType(),
     provider_id="basic",
     provider_resource_id="subset-of",
-    params=BasicScoringFnParams(
-        aggregation_functions=[AggregationFunctionType.accuracy]
-    ),
+    params=BasicScoringFnParams(aggregation_functions=[AggregationFunctionType.accuracy]),
 )

llama_stack/providers/inline/scoring/basic/scoring_fn/regex_parser_scoring_fn.py

@@ -33,17 +33,14 @@ class RegexParserScoringFn(RegisteredBaseScoringFn):
         scoring_fn_identifier: Optional[str] = None,
         scoring_params: Optional[ScoringFnParams] = None,
     ) -> ScoringResultRow:
-        assert (
-            scoring_fn_identifier is not None
-        ), "Scoring function identifier not found."
+        assert scoring_fn_identifier is not None, "Scoring function identifier not found."
         fn_def = self.supported_fn_defs_registry[scoring_fn_identifier]
         if scoring_params is not None:
             fn_def.params = scoring_params
 
-        assert (
-            fn_def.params is not None
-            and fn_def.params.type == ScoringFnParamsType.regex_parser.value
-        ), f"RegexParserScoringFnParams not found for {fn_def}."
+        assert fn_def.params is not None and fn_def.params.type == ScoringFnParamsType.regex_parser.value, (
+            f"RegexParserScoringFnParams not found for {fn_def}."
+        )
 
         expected_answer = input_row["expected_answer"]
         generated_answer = input_row["generated_answer"]