arize-phoenix 8.10.0__tar.gz → 8.12.0__tar.gz

{arize_phoenix-8.10.0 → arize_phoenix-8.12.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: arize-phoenix
-Version: 8.10.0
+Version: 8.12.0
 Summary: AI Observability and Evaluation
 Project-URL: Documentation, https://docs.arize.com/phoenix/
 Project-URL: Issues, https://github.com/Arize-ai/phoenix/issues
@@ -55,7 +55,7 @@ Requires-Dist: uvicorn
 Requires-Dist: websockets
 Requires-Dist: wrapt>=1.17.2
 Provides-Extra: container
-Requires-Dist: anthropic>=0.47.0; extra == 'container'
+Requires-Dist: anthropic>=0.49.0; extra == 'container'
 Requires-Dist: fast-hdbscan>=0.2.0; extra == 'container'
 Requires-Dist: google-generativeai; extra == 'container'
 Requires-Dist: numba>=0.60.0; extra == 'container'
@@ -73,7 +73,7 @@ Requires-Dist: strawberry-graphql[opentelemetry]==0.262.0; extra == 'container'
 Requires-Dist: umap-learn; extra == 'container'
 Requires-Dist: uvloop; (platform_system != 'Windows') and extra == 'container'
 Provides-Extra: dev
-Requires-Dist: anthropic>=0.47.0; extra == 'dev'
+Requires-Dist: anthropic>=0.49.0; extra == 'dev'
 Requires-Dist: arize[autoembeddings,llm-evaluation]; extra == 'dev'
 Requires-Dist: asgi-lifespan; extra == 'dev'
 Requires-Dist: asyncpg; extra == 'dev'

{arize_phoenix-8.10.0 → arize_phoenix-8.12.0}/pyproject.toml

@@ -95,7 +95,7 @@ dev = [
   "langchain>=0.0.334",
   "litellm>=1.0.3,<1.57.5", # windows compatibility broken on 1.57.5 (https://github.com/BerriAI/litellm/issues/7677)
   "google-cloud-aiplatform>=1.3",
-  "anthropic>=0.47.0",
+  "anthropic>=0.49.0",
   "prometheus_client",
   "asgi-lifespan",
   "Faker>=30.1.0",
@@ -123,7 +123,7 @@ pg = [
   "psycopg[binary,pool]",
 ]
 container = [
-  "anthropic>=0.47.0",
+  "anthropic>=0.49.0",
   "google-generativeai",
   "prometheus-client",
   "openai>=1.0.0",

{arize_phoenix-8.10.0 → arize_phoenix-8.12.0}/src/phoenix/db/models.py

@@ -970,6 +970,9 @@ class ExperimentRun(Base):
         primaryjoin="foreign(ExperimentRun.trace_id) == Trace.trace_id",
         back_populates="experiment_runs",
     )
+    annotations: Mapped[list["ExperimentRunAnnotation"]] = relationship(
+        back_populates="experiment_run"
+    )
 
     __table_args__ = (
         UniqueConstraint(
@@ -999,6 +1002,10 @@ class ExperimentRunAnnotation(Base):
     start_time: Mapped[datetime] = mapped_column(UtcTimeStamp)
     end_time: Mapped[datetime] = mapped_column(UtcTimeStamp)
 
+    experiment_run: Mapped["ExperimentRun"] = relationship(
+        back_populates="annotations",
+    )
+
     __table_args__ = (
         UniqueConstraint(
             "experiment_run_id",

{arize_phoenix-8.10.0 → arize_phoenix-8.12.0}/src/phoenix/server/api/helpers/playground_clients.py

@@ -761,13 +761,6 @@ class AnthropicStreamingClient(PlaygroundStreamingClient):
         import anthropic.types as anthropic_types
 
         anthropic_messages, system_prompt = self._build_anthropic_messages(messages)
-        if (
-            "thinking" in invocation_parameters
-            and isinstance(invocation_parameters, dict)
-            and "type" in invocation_parameters["thinking"]
-            and invocation_parameters["thinking"]["type"] == "enabled"
-        ):
-            invocation_parameters.pop("top_p", None)
         anthropic_params = {
             "messages": anthropic_messages,
             "model": self.model_name,

{arize_phoenix-8.10.0 → arize_phoenix-8.12.0}/src/phoenix/server/api/helpers/prompts/conversions/anthropic.py

@@ -13,6 +13,7 @@ if TYPE_CHECKING:
     )
 
     from phoenix.server.api.helpers.prompts.models import (
+        PromptToolChoiceNone,
         PromptToolChoiceOneOrMore,
         PromptToolChoiceSpecificFunctionTool,
         PromptToolChoiceZeroOrMore,
@@ -23,6 +24,7 @@ class AnthropicToolChoiceConversion:
     @staticmethod
     def to_anthropic(
         obj: Union[
+            PromptToolChoiceNone,
             PromptToolChoiceZeroOrMore,
             PromptToolChoiceOneOrMore,
             PromptToolChoiceSpecificFunctionTool,
@@ -44,6 +46,8 @@ class AnthropicToolChoiceConversion:
             if disable_parallel_tool_use is not None:
                 choice_tool["disable_parallel_tool_use"] = disable_parallel_tool_use
             return choice_tool
+        if obj.type == "none":
+            return {"type": "none"}
         assert_never(obj.type)
 
     @staticmethod
@@ -51,6 +55,7 @@ class AnthropicToolChoiceConversion:
         obj: ToolChoiceParam,
     ) -> tuple[
         Union[
+            PromptToolChoiceNone,
             PromptToolChoiceZeroOrMore,
             PromptToolChoiceOneOrMore,
             PromptToolChoiceSpecificFunctionTool,
@@ -58,6 +63,7 @@ class AnthropicToolChoiceConversion:
         Optional[bool],
     ]:
         from phoenix.server.api.helpers.prompts.models import (
+            PromptToolChoiceNone,
             PromptToolChoiceOneOrMore,
             PromptToolChoiceSpecificFunctionTool,
             PromptToolChoiceZeroOrMore,
@@ -84,4 +90,6 @@ class AnthropicToolChoiceConversion:
                 function_name=obj["name"],
             )
             return choice_function_tool, disable_parallel_tool_use
+        if obj["type"] == "none":
+            return PromptToolChoiceNone(type="none"), None
         assert_never(obj)

{arize_phoenix-8.10.0 → arize_phoenix-8.12.0}/src/phoenix/server/api/routers/v1/experiments.py

@@ -6,6 +6,7 @@ from typing import Any, Optional
 from fastapi import APIRouter, HTTPException, Path, Response
 from pydantic import Field
 from sqlalchemy import and_, func, select
+from sqlalchemy.orm import joinedload
 from starlette.requests import Request
 from starlette.responses import PlainTextResponse
 from starlette.status import HTTP_404_NOT_FOUND
@@ -321,7 +322,7 @@ async def list_experiments(
         ]
     ),
 )
-async def get_experiment_jsonl(
+async def get_experiment_json(
     request: Request,
     response: Response,
     experiment_id: str = Path(..., title="Experiment ID"),
@@ -359,27 +360,35 @@ async def get_experiment_jsonl(
             .scalar_subquery()
         )
         runs_and_revisions = (
-            await session.execute(
-                select(models.ExperimentRun, models.DatasetExampleRevision)
-                .join(
-                    models.DatasetExample,
-                    models.DatasetExample.id == models.ExperimentRun.dataset_example_id,
-                )
-                .join(
-                    models.DatasetExampleRevision,
-                    and_(
-                        models.DatasetExample.id
-                        == models.DatasetExampleRevision.dataset_example_id,
-                        models.DatasetExampleRevision.id.in_(revision_ids),
-                        models.DatasetExampleRevision.revision_kind != "DELETE",
-                    ),
-                )
-                .where(models.ExperimentRun.experiment_id == experiment_rowid)
-                .order_by(
-                    models.ExperimentRun.dataset_example_id, models.ExperimentRun.repetition_number
+            (
+                await session.execute(
+                    select(models.ExperimentRun, models.DatasetExampleRevision)
+                    .join(
+                        models.DatasetExample,
+                        models.DatasetExample.id == models.ExperimentRun.dataset_example_id,
+                    )
+                    .join(
+                        models.DatasetExampleRevision,
+                        and_(
+                            models.DatasetExample.id
+                            == models.DatasetExampleRevision.dataset_example_id,
+                            models.DatasetExampleRevision.id.in_(revision_ids),
+                            models.DatasetExampleRevision.revision_kind != "DELETE",
+                        ),
+                    )
+                    .options(
+                        joinedload(models.ExperimentRun.annotations),
+                    )
+                    .where(models.ExperimentRun.experiment_id == experiment_rowid)
+                    .order_by(
+                        models.ExperimentRun.dataset_example_id,
+                        models.ExperimentRun.repetition_number,
+                    )
                 )
             )
-        ).all()
+            .unique()
+            .all()
+        )
         if not runs_and_revisions:
             raise HTTPException(
                 detail=f"Experiment with ID {experiment_globalid} has no runs",
@@ -387,6 +396,22 @@ async def get_experiment_jsonl(
             )
     records = []
     for run, revision in runs_and_revisions:
+        annotations = []
+        for annotation in run.annotations:
+            annotations.append(
+                {
+                    "name": annotation.name,
+                    "annotator_kind": annotation.annotator_kind,
+                    "label": annotation.label,
+                    "score": annotation.score,
+                    "explanation": annotation.explanation,
+                    "trace_id": annotation.trace_id,
+                    "error": annotation.error,
+                    "metadata": annotation.metadata_,
+                    "start_time": annotation.start_time.isoformat(),
+                    "end_time": annotation.end_time.isoformat(),
+                }
+            )
         record = {
             "example_id": str(
                 GlobalID(models.DatasetExample.__name__, str(run.dataset_example_id))
@@ -394,7 +419,7 @@ async def get_experiment_jsonl(
             "repetition_number": run.repetition_number,
             "input": revision.input,
             "reference_output": revision.output,
-            "output": run.output,
+            "output": run.output["task_output"],
             "error": run.error,
             "latency_ms": run.latency_ms,
             "start_time": run.start_time.isoformat(),
@@ -402,6 +427,7 @@ async def get_experiment_jsonl(
             "trace_id": run.trace_id,
             "prompt_token_count": run.prompt_token_count,
             "completion_token_count": run.completion_token_count,
+            "annotations": annotations,
         }
         records.append(record)
 

arize_phoenix-8.12.0/src/phoenix/server/static/.vite/manifest.json

@@ -0,0 +1,87 @@
+{
+  "_components-CbkVj1OQ.js": {
+    "file": "assets/components-CbkVj1OQ.js",
+    "name": "components",
+    "imports": [
+      "_vendor-Byx9ZW8r.js",
+      "_pages-CyqzCvBY.js",
+      "_vendor-arizeai-0BoxbU9V.js",
+      "_vendor-codemirror-BX78TDyo.js",
+      "_vendor-three-C-AGeJYv.js"
+    ]
+  },
+  "_pages-CyqzCvBY.js": {
+    "file": "assets/pages-CyqzCvBY.js",
+    "name": "pages",
+    "imports": [
+      "_vendor-Byx9ZW8r.js",
+      "_vendor-arizeai-0BoxbU9V.js",
+      "_components-CbkVj1OQ.js",
+      "_vendor-codemirror-BX78TDyo.js",
+      "_vendor-recharts-BLborQJX.js"
+    ]
+  },
+  "_vendor-Byx9ZW8r.js": {
+    "file": "assets/vendor-Byx9ZW8r.js",
+    "name": "vendor",
+    "imports": [
+      "_vendor-three-C-AGeJYv.js"
+    ],
+    "css": [
+      "assets/vendor-Cg6lcjUC.css"
+    ]
+  },
+  "_vendor-Cg6lcjUC.css": {
+    "file": "assets/vendor-Cg6lcjUC.css",
+    "src": "_vendor-Cg6lcjUC.css"
+  },
+  "_vendor-arizeai-0BoxbU9V.js": {
+    "file": "assets/vendor-arizeai-0BoxbU9V.js",
+    "name": "vendor-arizeai",
+    "imports": [
+      "_vendor-Byx9ZW8r.js"
+    ]
+  },
+  "_vendor-codemirror-BX78TDyo.js": {
+    "file": "assets/vendor-codemirror-BX78TDyo.js",
+    "name": "vendor-codemirror",
+    "imports": [
+      "_vendor-Byx9ZW8r.js",
+      "_vendor-shiki-DpPpUgzv.js"
+    ]
+  },
+  "_vendor-recharts-BLborQJX.js": {
+    "file": "assets/vendor-recharts-BLborQJX.js",
+    "name": "vendor-recharts",
+    "imports": [
+      "_vendor-Byx9ZW8r.js"
+    ]
+  },
+  "_vendor-shiki-DpPpUgzv.js": {
+    "file": "assets/vendor-shiki-DpPpUgzv.js",
+    "name": "vendor-shiki",
+    "imports": [
+      "_vendor-Byx9ZW8r.js"
+    ]
+  },
+  "_vendor-three-C-AGeJYv.js": {
+    "file": "assets/vendor-three-C-AGeJYv.js",
+    "name": "vendor-three"
+  },
+  "index.tsx": {
+    "file": "assets/index-BNK6eNx3.js",
+    "name": "index",
+    "src": "index.tsx",
+    "isEntry": true,
+    "imports": [
+      "_vendor-Byx9ZW8r.js",
+      "_vendor-arizeai-0BoxbU9V.js",
+      "_pages-CyqzCvBY.js",
+      "_components-CbkVj1OQ.js",
+      "_vendor-three-C-AGeJYv.js",
+      "_vendor-codemirror-BX78TDyo.js",
+      "_vendor-shiki-DpPpUgzv.js",
+      "_vendor-recharts-BLborQJX.js"
+    ]
+  }
+}