arize-phoenix 4.21.0__py3-none-any.whl → 4.22.1__py3-none-any.whl

phoenix/server/api/context.py

@@ -14,7 +14,6 @@ from phoenix.server.api.dataloaders import (
     DocumentEvaluationsDataLoader,
     DocumentEvaluationSummaryDataLoader,
     DocumentRetrievalMetricsDataLoader,
-    EvaluationSummaryDataLoader,
     ExperimentAnnotationSummaryDataLoader,
     ExperimentErrorRatesDataLoader,
     ExperimentRunCountsDataLoader,
@@ -26,10 +25,8 @@ from phoenix.server.api.dataloaders import (
     SpanAnnotationsDataLoader,
     SpanDatasetExamplesDataLoader,
     SpanDescendantsDataLoader,
-    SpanEvaluationsDataLoader,
     SpanProjectsDataLoader,
     TokenCountDataLoader,
-    TraceEvaluationsDataLoader,
     TraceRowIdsDataLoader,
 )
 from phoenix.server.dml_event import DmlEvent
@@ -45,7 +42,6 @@ class DataLoaders:
     document_evaluations: DocumentEvaluationsDataLoader
     document_retrieval_metrics: DocumentRetrievalMetricsDataLoader
     annotation_summaries: AnnotationSummaryDataLoader
-    evaluation_summaries: EvaluationSummaryDataLoader
     experiment_annotation_summaries: ExperimentAnnotationSummaryDataLoader
     experiment_error_rates: ExperimentErrorRatesDataLoader
     experiment_run_counts: ExperimentRunCountsDataLoader
@@ -56,10 +52,8 @@ class DataLoaders:
     span_annotations: SpanAnnotationsDataLoader
     span_dataset_examples: SpanDatasetExamplesDataLoader
     span_descendants: SpanDescendantsDataLoader
-    span_evaluations: SpanEvaluationsDataLoader
     span_projects: SpanProjectsDataLoader
     token_counts: TokenCountDataLoader
-    trace_evaluations: TraceEvaluationsDataLoader
     trace_row_ids: TraceRowIdsDataLoader
     project_by_name: ProjectByNameDataLoader
 

phoenix/server/api/dataloaders/__init__.py

@@ -10,7 +10,6 @@ from .document_evaluation_summaries import (
 )
 from .document_evaluations import DocumentEvaluationsDataLoader
 from .document_retrieval_metrics import DocumentRetrievalMetricsDataLoader
-from .evaluation_summaries import EvaluationSummaryCache, EvaluationSummaryDataLoader
 from .experiment_annotation_summaries import ExperimentAnnotationSummaryDataLoader
 from .experiment_error_rates import ExperimentErrorRatesDataLoader
 from .experiment_run_counts import ExperimentRunCountsDataLoader
@@ -22,10 +21,8 @@ from .record_counts import RecordCountCache, RecordCountDataLoader
 from .span_annotations import SpanAnnotationsDataLoader
 from .span_dataset_examples import SpanDatasetExamplesDataLoader
 from .span_descendants import SpanDescendantsDataLoader
-from .span_evaluations import SpanEvaluationsDataLoader
 from .span_projects import SpanProjectsDataLoader
 from .token_counts import TokenCountCache, TokenCountDataLoader
-from .trace_evaluations import TraceEvaluationsDataLoader
 from .trace_row_ids import TraceRowIdsDataLoader
 
 __all__ = [
@@ -37,7 +34,6 @@ __all__ = [
     "DocumentEvaluationsDataLoader",
     "DocumentRetrievalMetricsDataLoader",
     "AnnotationSummaryDataLoader",
-    "EvaluationSummaryDataLoader",
     "ExperimentAnnotationSummaryDataLoader",
     "ExperimentErrorRatesDataLoader",
     "ExperimentRunCountsDataLoader",
@@ -47,10 +43,8 @@ __all__ = [
     "RecordCountDataLoader",
     "SpanDatasetExamplesDataLoader",
     "SpanDescendantsDataLoader",
-    "SpanEvaluationsDataLoader",
     "SpanProjectsDataLoader",
     "TokenCountDataLoader",
-    "TraceEvaluationsDataLoader",
     "TraceRowIdsDataLoader",
     "ProjectByNameDataLoader",
     "SpanAnnotationsDataLoader",
@@ -65,9 +59,6 @@ class CacheForDataLoaders:
     annotation_summary: AnnotationSummaryCache = field(
         default_factory=AnnotationSummaryCache,
     )
-    evaluation_summary: EvaluationSummaryCache = field(
-        default_factory=EvaluationSummaryCache,
-    )
     latency_ms_quantile: LatencyMsQuantileCache = field(
         default_factory=LatencyMsQuantileCache,
     )

phoenix/server/api/routers/v1/experiment_runs.py

@@ -108,7 +108,12 @@ async def create_experiment_run(
     return CreateExperimentResponseBody(data=CreateExperimentRunResponseBodyData(id=str(run_gid)))
 
 
-class ListExperimentRunsResponseBody(ResponseBody[List[ExperimentRun]]):
+class ExperimentRunResponse(ExperimentRun):
+    id: str = Field(description="The ID of the experiment run")
+    experiment_id: str = Field(description="The ID of the experiment")
+
+
+class ListExperimentRunsResponseBody(ResponseBody[List[ExperimentRunResponse]]):
     pass
 
 
@@ -147,7 +152,7 @@ async def list_experiment_runs(
             experiment_gid = GlobalID("Experiment", str(exp_run.experiment_id))
             example_gid = GlobalID("DatasetExample", str(exp_run.dataset_example_id))
             runs.append(
-                ExperimentRun(
+                ExperimentRunResponse(
                     start_time=exp_run.start_time,
                     end_time=exp_run.end_time,
                     experiment_id=str(experiment_gid),

phoenix/server/api/types/Evaluation.py

@@ -1,7 +1,7 @@
 import strawberry
 
 import phoenix.trace.v1 as pb
-from phoenix.db.models import DocumentAnnotation, SpanAnnotation, TraceAnnotation
+from phoenix.db.models import DocumentAnnotation, TraceAnnotation
 
 from .Annotation import Annotation
 
@@ -31,31 +31,6 @@ class TraceEvaluation(Annotation):
         )
 
 
-@strawberry.type
-class SpanEvaluation(Annotation):
-    @staticmethod
-    def from_pb_evaluation(evaluation: pb.Evaluation) -> "SpanEvaluation":
-        result = evaluation.result
-        score = result.score.value if result.HasField("score") else None
-        label = result.label.value if result.HasField("label") else None
-        explanation = result.explanation.value if result.HasField("explanation") else None
-        return SpanEvaluation(
-            name=evaluation.name,
-            score=score,
-            label=label,
-            explanation=explanation,
-        )
-
-    @staticmethod
-    def from_sql_span_annotation(annotation: SpanAnnotation) -> "SpanEvaluation":
-        return SpanEvaluation(
-            name=annotation.name,
-            score=annotation.score,
-            label=annotation.label,
-            explanation=annotation.explanation,
-        )
-
-
 @strawberry.type
 class DocumentEvaluation(Annotation):
     document_position: int = strawberry.field(

phoenix/server/api/types/Project.py

@@ -24,7 +24,6 @@ from phoenix.server.api.input_types.SpanSort import SpanSort, SpanSortConfig
 from phoenix.server.api.input_types.TimeRange import TimeRange
 from phoenix.server.api.types.AnnotationSummary import AnnotationSummary
 from phoenix.server.api.types.DocumentEvaluationSummary import DocumentEvaluationSummary
-from phoenix.server.api.types.EvaluationSummary import EvaluationSummary
 from phoenix.server.api.types.pagination import (
     Cursor,
     CursorSortColumn,
@@ -249,23 +248,6 @@ class Project(Node):
             has_next_page=has_next_page,
         )
 
-    @strawberry.field(
-        description="Names of all available evaluations for traces. "
-        "(The list contains no duplicates.)"
-    )  # type: ignore
-    async def trace_evaluation_names(
-        self,
-        info: Info[Context, None],
-    ) -> List[str]:
-        stmt = (
-            select(distinct(models.TraceAnnotation.name))
-            .join(models.Trace)
-            .where(models.Trace.project_rowid == self.id_attr)
-            .where(models.TraceAnnotation.annotator_kind == "LLM")
-        )
-        async with info.context.db() as session:
-            return list(await session.scalars(stmt))
-
     @strawberry.field(
         description="Names of all available annotations for traces. "
         "(The list contains no duplicates.)"
@@ -282,24 +264,6 @@ class Project(Node):
         async with info.context.db() as session:
             return list(await session.scalars(stmt))
 
-    @strawberry.field(
-        description="Names of all available evaluations for spans. "
-        "(The list contains no duplicates.)"
-    )  # type: ignore
-    async def span_evaluation_names(
-        self,
-        info: Info[Context, None],
-    ) -> List[str]:
-        stmt = (
-            select(distinct(models.SpanAnnotation.name))
-            .join(models.Span)
-            .join(models.Trace, models.Span.trace_rowid == models.Trace.id)
-            .where(models.Trace.project_rowid == self.id_attr)
-            .where(models.SpanAnnotation.annotator_kind == "LLM")
-        )
-        async with info.context.db() as session:
-            return list(await session.scalars(stmt))
-
     @strawberry.field(
         description="Names of all available annotations for spans. "
         "(The list contains no duplicates.)"
@@ -337,29 +301,6 @@ class Project(Node):
         async with info.context.db() as session:
             return list(await session.scalars(stmt))
 
-    @strawberry.field
-    async def trace_evaluation_summary(
-        self,
-        info: Info[Context, None],
-        evaluation_name: str,
-        time_range: Optional[TimeRange] = UNSET,
-    ) -> Optional[EvaluationSummary]:
-        return await info.context.data_loaders.evaluation_summaries.load(
-            ("trace", self.id_attr, time_range, None, evaluation_name),
-        )
-
-    @strawberry.field
-    async def span_evaluation_summary(
-        self,
-        info: Info[Context, None],
-        evaluation_name: str,
-        time_range: Optional[TimeRange] = UNSET,
-        filter_condition: Optional[str] = UNSET,
-    ) -> Optional[EvaluationSummary]:
-        return await info.context.data_loaders.evaluation_summaries.load(
-            ("span", self.id_attr, time_range, filter_condition, evaluation_name),
-        )
-
     @strawberry.field
     async def trace_annotation_summary(
         self,
@@ -405,7 +346,7 @@ class Project(Node):
     @strawberry.field
     async def validate_span_filter_condition(self, condition: str) -> ValidationResult:
         # This query is too expensive to run on every validation
-        # valid_eval_names = await self.span_evaluation_names()
+        # valid_eval_names = await self.span_annotation_names()
         try:
             SpanFilter(
                 condition=condition,

phoenix/server/api/types/Span.py

@@ -28,7 +28,7 @@ from phoenix.server.api.types.SpanAnnotation import to_gql_span_annotation
 from phoenix.trace.attributes import get_attribute_value
 
 from .DocumentRetrievalMetrics import DocumentRetrievalMetrics
-from .Evaluation import DocumentEvaluation, SpanEvaluation
+from .Evaluation import DocumentEvaluation
 from .ExampleRevisionInterface import ExampleRevision
 from .MimeType import MimeType
 from .SpanAnnotation import SpanAnnotation
@@ -170,14 +170,6 @@ class Span(Node):
         "codes from descendant spans (children, grandchildren, etc.)",
     )
 
-    @strawberry.field(
-        description="Evaluations associated with the span, e.g. if the span is "
-        "an LLM, an evaluation may assess the helpfulness of its response with "
-        "respect to its input."
-    )  # type: ignore
-    async def span_evaluations(self, info: Info[Context, None]) -> List[SpanEvaluation]:
-        return await info.context.data_loaders.span_evaluations.load(self.id_attr)
-
     @strawberry.field(
         description=(
             "Annotations associated with the span. This encompasses both "

phoenix/server/app.py

@@ -65,7 +65,6 @@ from phoenix.server.api.dataloaders import (
     DocumentEvaluationsDataLoader,
     DocumentEvaluationSummaryDataLoader,
     DocumentRetrievalMetricsDataLoader,
-    EvaluationSummaryDataLoader,
     ExperimentAnnotationSummaryDataLoader,
     ExperimentErrorRatesDataLoader,
     ExperimentRunCountsDataLoader,
@@ -77,10 +76,8 @@ from phoenix.server.api.dataloaders import (
     SpanAnnotationsDataLoader,
     SpanDatasetExamplesDataLoader,
     SpanDescendantsDataLoader,
-    SpanEvaluationsDataLoader,
     SpanProjectsDataLoader,
     TokenCountDataLoader,
-    TraceEvaluationsDataLoader,
     TraceRowIdsDataLoader,
 )
 from phoenix.server.api.routers.v1 import REST_API_VERSION
@@ -303,12 +300,6 @@ def create_graphql_router(
                     if cache_for_dataloaders
                     else None,
                 ),
-                evaluation_summaries=EvaluationSummaryDataLoader(
-                    db,
-                    cache_map=cache_for_dataloaders.evaluation_summary
-                    if cache_for_dataloaders
-                    else None,
-                ),
                 experiment_annotation_summaries=ExperimentAnnotationSummaryDataLoader(db),
                 experiment_error_rates=ExperimentErrorRatesDataLoader(db),
                 experiment_run_counts=ExperimentRunCountsDataLoader(db),
@@ -332,13 +323,11 @@ def create_graphql_router(
                 span_annotations=SpanAnnotationsDataLoader(db),
                 span_dataset_examples=SpanDatasetExamplesDataLoader(db),
                 span_descendants=SpanDescendantsDataLoader(db),
-                span_evaluations=SpanEvaluationsDataLoader(db),
                 span_projects=SpanProjectsDataLoader(db),
                 token_counts=TokenCountDataLoader(
                     db,
                     cache_map=cache_for_dataloaders.token_count if cache_for_dataloaders else None,
                 ),
-                trace_evaluations=TraceEvaluationsDataLoader(db),
                 trace_row_ids=TraceRowIdsDataLoader(db),
                 project_by_name=ProjectByNameDataLoader(db),
             ),

phoenix/server/dml_event_handler.py

@@ -142,7 +142,6 @@ class _SpanDeleteEventHandler(_SpanDmlEventHandler):
     @staticmethod
     def _clear(cache: CacheForDataLoaders, project_id: int) -> None:
         cache.annotation_summary.invalidate_project(project_id)
-        cache.evaluation_summary.invalidate_project(project_id)
         cache.document_evaluation_summary.invalidate_project(project_id)
 
 
@@ -201,7 +200,6 @@ class _SpanAnnotationDmlEventHandler(_AnnotationDmlEventHandler[SpanAnnotationDm
     @staticmethod
     def _clear(cache: CacheForDataLoaders, project_id: int, name: str) -> None:
         cache.annotation_summary.invalidate((project_id, name, "span"))
-        cache.evaluation_summary.invalidate((project_id, name, "span"))
 
 
 class _TraceAnnotationDmlEventHandler(_AnnotationDmlEventHandler[TraceAnnotationDmlEvent]):
@@ -214,7 +212,6 @@ class _TraceAnnotationDmlEventHandler(_AnnotationDmlEventHandler[TraceAnnotation
     @staticmethod
     def _clear(cache: CacheForDataLoaders, project_id: int, name: str) -> None:
         cache.annotation_summary.invalidate((project_id, name, "trace"))
-        cache.evaluation_summary.invalidate((project_id, name, "trace"))
 
 
 class _DocumentAnnotationDmlEventHandler(_AnnotationDmlEventHandler[DocumentAnnotationDmlEvent]):

phoenix/server/static/.vite/manifest.json

@@ -1,21 +1,21 @@
 {
-  "_components-D2V-mOGq.js": {
-    "file": "assets/components-D2V-mOGq.js",
+  "_components-BC3-LP_a.js": {
+    "file": "assets/components-BC3-LP_a.js",
     "name": "components",
     "imports": [
       "_vendor-BMWfu6zp.js",
       "_vendor-arizeai-Sj74jm5V.js",
-      "_pages-CChOjmat.js",
+      "_pages--n2933VW.js",
       "_vendor-three-DwGkEfCM.js",
       "_vendor-codemirror-DO3VqEcD.js"
     ]
   },
-  "_pages-CChOjmat.js": {
-    "file": "assets/pages-CChOjmat.js",
+  "_pages--n2933VW.js": {
+    "file": "assets/pages--n2933VW.js",
     "name": "pages",
     "imports": [
       "_vendor-BMWfu6zp.js",
-      "_components-D2V-mOGq.js",
+      "_components-BC3-LP_a.js",
       "_vendor-arizeai-Sj74jm5V.js",
       "_vendor-recharts-BGN0SxgJ.js",
       "_vendor-codemirror-DO3VqEcD.js"
@@ -61,15 +61,15 @@
     "name": "vendor-three"
   },
   "index.tsx": {
-    "file": "assets/index-B52Z3aZG.js",
+    "file": "assets/index-BjJvafYL.js",
     "name": "index",
     "src": "index.tsx",
     "isEntry": true,
     "imports": [
       "_vendor-BMWfu6zp.js",
       "_vendor-arizeai-Sj74jm5V.js",
-      "_components-D2V-mOGq.js",
-      "_pages-CChOjmat.js",
+      "_components-BC3-LP_a.js",
+      "_pages--n2933VW.js",
       "_vendor-three-DwGkEfCM.js",
       "_vendor-codemirror-DO3VqEcD.js",
       "_vendor-recharts-BGN0SxgJ.js"