pydantic-evals 1.1.0__tar.gz → 1.10.0__tar.gz

{pydantic_evals-1.1.0 → pydantic_evals-1.10.0}/.gitignore

@@ -10,7 +10,7 @@ env*/
 /TODO.md
 /postgres-data/
 .DS_Store
-examples/pydantic_ai_examples/.chat_app_messages.sqlite
+.chat_app_messages.sqlite
 .cache/
 .vscode/
 /question_graph_history.json

{pydantic_evals-1.1.0 → pydantic_evals-1.10.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pydantic-evals
-Version: 1.1.0
+Version: 1.10.0
 Summary: Framework for evaluating stochastic code execution, especially code making use of LLMs
 Project-URL: Homepage, https://ai.pydantic.dev/evals
 Project-URL: Source, https://github.com/pydantic/pydantic-ai
@@ -30,7 +30,7 @@ Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.10
 Requires-Dist: anyio>=0
 Requires-Dist: logfire-api>=3.14.1
-Requires-Dist: pydantic-ai-slim==1.1.0
+Requires-Dist: pydantic-ai-slim==1.10.0
 Requires-Dist: pydantic>=2.10
 Requires-Dist: pyyaml>=6.0.2
 Requires-Dist: rich>=13.9.4

pydantic_evals-1.10.0/pydantic_evals/__init__.py

@@ -0,0 +1,16 @@
+"""A toolkit for evaluating the execution of arbitrary "stochastic functions", such as LLM calls.
+
+This package provides functionality for:
+- Creating and loading test datasets with structured inputs and outputs
+- Evaluating model performance using various metrics and evaluators
+- Generating reports for evaluation results
+"""
+
+from .dataset import Case, Dataset, increment_eval_metric, set_eval_attribute
+
+__all__ = (
+    'Case',
+    'Dataset',
+    'increment_eval_metric',
+    'set_eval_attribute',
+)

{pydantic_evals-1.1.0 → pydantic_evals-1.10.0}/pydantic_evals/dataset.py

@@ -265,6 +265,7 @@ class Dataset(BaseModel, Generic[InputsT, OutputT, MetadataT], extra='forbid', a
         retry_evaluators: RetryConfig | None = None,
         *,
         task_name: str | None = None,
+        metadata: dict[str, Any] | None = None,
     ) -> EvaluationReport[InputsT, OutputT, MetadataT]:
         """Evaluates the test cases in the dataset using the given task.
 
@@ -283,6 +284,7 @@ class Dataset(BaseModel, Generic[InputsT, OutputT, MetadataT], extra='forbid', a
             retry_evaluators: Optional retry configuration for evaluator execution.
             task_name: Optional override to the name of the task being executed, otherwise the name of the task
                 function will be used.
+            metadata: Optional dict of experiment metadata.
 
         Returns:
             A report containing the results of the evaluation.
@@ -294,6 +296,9 @@ class Dataset(BaseModel, Generic[InputsT, OutputT, MetadataT], extra='forbid', a
 
         limiter = anyio.Semaphore(max_concurrency) if max_concurrency is not None else AsyncExitStack()
 
+        extra_attributes: dict[str, Any] = {'gen_ai.operation.name': 'experiment'}
+        if metadata is not None:
+            extra_attributes['metadata'] = metadata
         with (
             logfire_span(
                 'evaluate {name}',
@@ -301,7 +306,7 @@ class Dataset(BaseModel, Generic[InputsT, OutputT, MetadataT], extra='forbid', a
                 task_name=task_name,
                 dataset_name=self.name,
                 n_cases=len(self.cases),
-                **{'gen_ai.operation.name': 'experiment'},  # pyright: ignore[reportArgumentType]
+                **extra_attributes,
             ) as eval_span,
             progress_bar or nullcontext(),
         ):
@@ -339,11 +344,18 @@ class Dataset(BaseModel, Generic[InputsT, OutputT, MetadataT], extra='forbid', a
                 name=name,
                 cases=cases,
                 failures=failures,
+                experiment_metadata=metadata,
                 span_id=span_id,
                 trace_id=trace_id,
             )
-            if (averages := report.averages()) is not None and averages.assertions is not None:
-                eval_span.set_attribute('assertion_pass_rate', averages.assertions)
+            full_experiment_metadata: dict[str, Any] = {'n_cases': len(self.cases)}
+            if metadata is not None:
+                full_experiment_metadata['metadata'] = metadata
+            if (averages := report.averages()) is not None:
+                full_experiment_metadata['averages'] = averages
+                if averages.assertions is not None:
+                    eval_span.set_attribute('assertion_pass_rate', averages.assertions)
+            eval_span.set_attribute('logfire.experiment.metadata', full_experiment_metadata)
         return report
 
     def evaluate_sync(
@@ -354,6 +366,9 @@ class Dataset(BaseModel, Generic[InputsT, OutputT, MetadataT], extra='forbid', a
         progress: bool = True,
         retry_task: RetryConfig | None = None,
         retry_evaluators: RetryConfig | None = None,
+        *,
+        task_name: str | None = None,
+        metadata: dict[str, Any] | None = None,
     ) -> EvaluationReport[InputsT, OutputT, MetadataT]:
         """Evaluates the test cases in the dataset using the given task.
 
@@ -362,13 +377,16 @@ class Dataset(BaseModel, Generic[InputsT, OutputT, MetadataT], extra='forbid', a
         Args:
             task: The task to evaluate. This should be a callable that takes the inputs of the case
                 and returns the output.
-            name: The name of the task being evaluated, this is used to identify the task in the report.
-                If omitted, the name of the task function will be used.
+            name: The name of the experiment being run, this is used to identify the experiment in the report.
+                If omitted, the task_name will be used; if that is not specified, the name of the task function is used.
             max_concurrency: The maximum number of concurrent evaluations of the task to allow.
                 If None, all cases will be evaluated concurrently.
-            progress: Whether to show a progress bar for the evaluation. Defaults to True.
+            progress: Whether to show a progress bar for the evaluation. Defaults to `True`.
             retry_task: Optional retry configuration for the task execution.
             retry_evaluators: Optional retry configuration for evaluator execution.
+            task_name: Optional override to the name of the task being executed, otherwise the name of the task
+                function will be used.
+            metadata: Optional dict of experiment metadata.
 
         Returns:
             A report containing the results of the evaluation.
@@ -376,11 +394,13 @@ class Dataset(BaseModel, Generic[InputsT, OutputT, MetadataT], extra='forbid', a
         return get_event_loop().run_until_complete(
             self.evaluate(
                 task,
-                task_name=name,
+                name=name,
                 max_concurrency=max_concurrency,
                 progress=progress,
                 retry_task=retry_task,
                 retry_evaluators=retry_evaluators,
+                task_name=task_name,
+                metadata=metadata,
             )
         )
 

{pydantic_evals-1.1.0 → pydantic_evals-1.10.0}/pydantic_evals/evaluators/llm_as_a_judge.py

@@ -201,7 +201,7 @@ async def judge_output_expected(
     ).output
 
 
-def set_default_judge_model(model: models.Model | models.KnownModelName) -> None:  # pragma: no cover
+def set_default_judge_model(model: models.Model | models.KnownModelName) -> None:
     """Set the default model used for judging.
 
     This model is used if `None` is passed to the `model` argument of `judge_output` and `judge_input_output`.

{pydantic_evals-1.1.0 → pydantic_evals-1.10.0}/pydantic_evals/generation.py

@@ -14,6 +14,7 @@ from pydantic import ValidationError
 from typing_extensions import TypeVar
 
 from pydantic_ai import Agent, models
+from pydantic_ai._utils import strip_markdown_fences
 from pydantic_evals import Dataset
 from pydantic_evals.evaluators.evaluator import Evaluator
 
@@ -73,8 +74,9 @@ async def generate_dataset(
     )
 
     result = await agent.run(extra_instructions or 'Please generate the object.')
+    output = strip_markdown_fences(result.output)
     try:
-        result = dataset_type.from_text(result.output, fmt='json', custom_evaluator_types=custom_evaluator_types)
+        result = dataset_type.from_text(output, fmt='json', custom_evaluator_types=custom_evaluator_types)
     except ValidationError as e:  # pragma: no cover
         print(f'Raw response from model:\n{result.output}')
         raise e

{pydantic_evals-1.1.0 → pydantic_evals-1.10.0}/pydantic_evals/reporting/__init__.py

@@ -7,8 +7,10 @@ from io import StringIO
 from typing import Any, Generic, Literal, Protocol, cast
 
 from pydantic import BaseModel, TypeAdapter
-from rich.console import Console
+from rich.console import Console, Group, RenderableType
+from rich.panel import Panel
 from rich.table import Table
+from rich.text import Text
 from typing_extensions import TypedDict, TypeVar
 
 from pydantic_evals._utils import UNSET, Unset
@@ -196,6 +198,8 @@ class EvaluationReport(Generic[InputsT, OutputT, MetadataT]):
     failures: list[ReportCaseFailure[InputsT, OutputT, MetadataT]] = field(default_factory=list)
     """The failures in the report. These are cases where task execution raised an exception."""
 
+    experiment_metadata: dict[str, Any] | None = None
+    """Metadata associated with the specific experiment represented by this report."""
     trace_id: str | None = None
     """The trace ID of the evaluation."""
     span_id: str | None = None
@@ -230,7 +234,7 @@ class EvaluationReport(Generic[InputsT, OutputT, MetadataT]):
         metric_configs: dict[str, RenderNumberConfig] | None = None,
         duration_config: RenderNumberConfig | None = None,
         include_reasons: bool = False,
-    ) -> str:  # pragma: no cover
+    ) -> str:
         """Render this report to a nicely-formatted string, optionally comparing it to a baseline report.
 
         If you want more control over the output, use `console_table` instead and pass it to `rich.Console.print`.
@@ -261,7 +265,6 @@ class EvaluationReport(Generic[InputsT, OutputT, MetadataT]):
             duration_config=duration_config,
             include_reasons=include_reasons,
         )
-        Console(file=io_file)
         return io_file.getvalue()
 
     def print(
@@ -289,15 +292,16 @@ class EvaluationReport(Generic[InputsT, OutputT, MetadataT]):
         metric_configs: dict[str, RenderNumberConfig] | None = None,
         duration_config: RenderNumberConfig | None = None,
         include_reasons: bool = False,
-    ) -> None:  # pragma: no cover
+    ) -> None:
         """Print this report to the console, optionally comparing it to a baseline report.
 
         If you want more control over the output, use `console_table` instead and pass it to `rich.Console.print`.
         """
-        if console is None:
+        if console is None:  # pragma: no branch
             console = Console(width=width)
 
-        table = self.console_table(
+        metadata_panel = self._metadata_panel(baseline=baseline)
+        renderable: RenderableType = self.console_table(
             baseline=baseline,
             include_input=include_input,
             include_metadata=include_metadata,
@@ -316,9 +320,13 @@ class EvaluationReport(Generic[InputsT, OutputT, MetadataT]):
             metric_configs=metric_configs,
             duration_config=duration_config,
             include_reasons=include_reasons,
+            with_title=not metadata_panel,
         )
-        console.print(table)
-        if include_errors and self.failures:
+        # Wrap table with experiment metadata panel if present
+        if metadata_panel:
+            renderable = Group(metadata_panel, renderable)
+        console.print(renderable)
+        if include_errors and self.failures:  # pragma: no cover
             failures_table = self.failures_table(
                 include_input=include_input,
                 include_metadata=include_metadata,
@@ -330,6 +338,7 @@ class EvaluationReport(Generic[InputsT, OutputT, MetadataT]):
             )
             console.print(failures_table, style='red')
 
+    # TODO(DavidM): in v2, change the return type here to RenderableType
     def console_table(
         self,
         baseline: EvaluationReport[InputsT, OutputT, MetadataT] | None = None,
@@ -351,9 +360,11 @@ class EvaluationReport(Generic[InputsT, OutputT, MetadataT]):
         metric_configs: dict[str, RenderNumberConfig] | None = None,
         duration_config: RenderNumberConfig | None = None,
         include_reasons: bool = False,
+        with_title: bool = True,
     ) -> Table:
-        """Return a table containing the data from this report, or the diff between this report and a baseline report.
+        """Return a table containing the data from this report.
 
+        If a baseline is provided, returns a diff between this report and the baseline report.
         Optionally include input and output details.
         """
         renderer = EvaluationRenderer(
@@ -378,10 +389,82 @@ class EvaluationReport(Generic[InputsT, OutputT, MetadataT]):
             include_reasons=include_reasons,
         )
         if baseline is None:
-            return renderer.build_table(self)
-        else:  # pragma: no cover
-            return renderer.build_diff_table(self, baseline)
+            return renderer.build_table(self, with_title=with_title)
+        else:
+            return renderer.build_diff_table(self, baseline, with_title=with_title)
+
+    def _metadata_panel(
+        self, baseline: EvaluationReport[InputsT, OutputT, MetadataT] | None = None
+    ) -> RenderableType | None:
+        """Wrap a table with an experiment metadata panel if metadata exists.
 
+        Args:
+            table: The table to wrap
+            baseline: Optional baseline report for diff metadata
+
+        Returns:
+            Either the table unchanged or a Group with Panel and Table
+        """
+        if baseline is None:
+            # Single report - show metadata if present
+            if self.experiment_metadata:
+                metadata_text = Text()
+                items = list(self.experiment_metadata.items())
+                for i, (key, value) in enumerate(items):
+                    metadata_text.append(f'{key}: {value}', style='dim')
+                    if i < len(items) - 1:
+                        metadata_text.append('\n')
+                return Panel(
+                    metadata_text,
+                    title=f'Evaluation Summary: {self.name}',
+                    title_align='left',
+                    border_style='dim',
+                    padding=(0, 1),
+                    expand=False,
+                )
+        else:
+            # Diff report - show metadata diff if either has metadata
+            if self.experiment_metadata or baseline.experiment_metadata:
+                diff_name = baseline.name if baseline.name == self.name else f'{baseline.name} → {self.name}'
+                metadata_text = Text()
+                lines_styles: list[tuple[str, str]] = []
+                if baseline.experiment_metadata and self.experiment_metadata:
+                    # Collect all keys from both
+                    all_keys = sorted(set(baseline.experiment_metadata.keys()) | set(self.experiment_metadata.keys()))
+                    for key in all_keys:
+                        baseline_val = baseline.experiment_metadata.get(key)
+                        report_val = self.experiment_metadata.get(key)
+                        if baseline_val == report_val:
+                            lines_styles.append((f'{key}: {report_val}', 'dim'))
+                        elif baseline_val is None:
+                            lines_styles.append((f'+ {key}: {report_val}', 'green'))
+                        elif report_val is None:
+                            lines_styles.append((f'- {key}: {baseline_val}', 'red'))
+                        else:
+                            lines_styles.append((f'{key}: {baseline_val} → {report_val}', 'yellow'))
+                elif self.experiment_metadata:
+                    lines_styles = [(f'+ {k}: {v}', 'green') for k, v in self.experiment_metadata.items()]
+                else:  # baseline.experiment_metadata only
+                    assert baseline.experiment_metadata is not None
+                    lines_styles = [(f'- {k}: {v}', 'red') for k, v in baseline.experiment_metadata.items()]
+
+                for i, (line, style) in enumerate(lines_styles):
+                    metadata_text.append(line, style=style)
+                    if i < len(lines_styles) - 1:
+                        metadata_text.append('\n')
+
+                return Panel(
+                    metadata_text,
+                    title=f'Evaluation Diff: {diff_name}',
+                    title_align='left',
+                    border_style='dim',
+                    padding=(0, 1),
+                    expand=False,
+                )
+
+        return None
+
+    # TODO(DavidM): in v2, change the return type here to RenderableType
     def failures_table(
         self,
         *,
@@ -705,6 +788,7 @@ class ReportCaseRenderer:
     metric_renderers: Mapping[str, _NumberRenderer]
     duration_renderer: _NumberRenderer
 
+    # TODO(DavidM): in v2, change the return type here to RenderableType
     def build_base_table(self, title: str) -> Table:
         """Build and return a Rich Table for the diff output."""
         table = Table(title=title, show_lines=True)
@@ -731,6 +815,7 @@ class ReportCaseRenderer:
             table.add_column('Durations' if self.include_total_duration else 'Duration', justify='right')
         return table
 
+    # TODO(DavidM): in v2, change the return type here to RenderableType
     def build_failures_table(self, title: str) -> Table:
         """Build and return a Rich Table for the failures output."""
         table = Table(title=title, show_lines=True)
@@ -1190,9 +1275,22 @@ class EvaluationRenderer:
             duration_renderer=duration_renderer,
         )
 
-    def build_table(self, report: EvaluationReport) -> Table:
+    # TODO(DavidM): in v2, change the return type here to RenderableType
+    def build_table(self, report: EvaluationReport, *, with_title: bool = True) -> Table:
+        """Build a table for the report.
+
+        Args:
+            report: The evaluation report to render
+            with_title: Whether to include the title in the table (default True)
+
+        Returns:
+            A Rich Table object
+        """
         case_renderer = self._get_case_renderer(report)
-        table = case_renderer.build_base_table(f'Evaluation Summary: {report.name}')
+
+        title = f'Evaluation Summary: {report.name}' if with_title else ''
+        table = case_renderer.build_base_table(title)
+
         for case in report.cases:
             table.add_row(*case_renderer.build_row(case))
 
@@ -1203,7 +1301,20 @@ class EvaluationRenderer:
 
         return table
 
-    def build_diff_table(self, report: EvaluationReport, baseline: EvaluationReport) -> Table:
+    # TODO(DavidM): in v2, change the return type here to RenderableType
+    def build_diff_table(
+        self, report: EvaluationReport, baseline: EvaluationReport, *, with_title: bool = True
+    ) -> Table:
+        """Build a diff table comparing report to baseline.
+
+        Args:
+            report: The evaluation report to compare
+            baseline: The baseline report to compare against
+            with_title: Whether to include the title in the table (default True)
+
+        Returns:
+            A Rich Table object
+        """
         report_cases = report.cases
         baseline_cases = self._baseline_cases_to_include(report, baseline)
 
@@ -1228,7 +1339,10 @@ class EvaluationRenderer:
 
         case_renderer = self._get_case_renderer(report, baseline)
         diff_name = baseline.name if baseline.name == report.name else f'{baseline.name} → {report.name}'
-        table = case_renderer.build_base_table(f'Evaluation Diff: {diff_name}')
+
+        title = f'Evaluation Diff: {diff_name}' if with_title else ''
+        table = case_renderer.build_base_table(title)
+
         for baseline_case, new_case in diff_cases:
             table.add_row(*case_renderer.build_diff_row(new_case, baseline_case))
         for case in added_cases:
@@ -1247,6 +1361,7 @@ class EvaluationRenderer:
 
         return table
 
+    # TODO(DavidM): in v2, change the return type here to RenderableType
     def build_failures_table(self, report: EvaluationReport) -> Table:
         case_renderer = self._get_case_renderer(report)
         table = case_renderer.build_failures_table('Case Failures')

pydantic_evals-1.1.0/pydantic_evals/__init__.py

@@ -1,19 +0,0 @@
-"""A toolkit for evaluating the execution of arbitrary "stochastic functions", such as LLM calls.
-
-This package provides functionality for:
-- Creating and loading test datasets with structured inputs and outputs
-- Evaluating model performance using various metrics and evaluators
-- Generating reports for evaluation results
-
-TODO(DavidM): Implement serialization of reports for later comparison, and add git hashes etc.
-  Note: I made pydantic_ai.evals.reports.EvalReport a BaseModel specifically to make this easier
-TODO(DavidM): Add commit hash, timestamp, and other metadata to reports (like pytest-speed does), possibly in a dedicated struct
-TODO(DavidM): Implement a CLI with some pytest-like filtering API to make it easier to run only specific cases
-"""
-
-from .dataset import Case, Dataset
-
-__all__ = (
-    'Case',
-    'Dataset',
-)