pydantic-evals 0.8.0__py3-none-any.whl → 0.8.1__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Potentially problematic release.


This version of pydantic-evals might be problematic. Click here for more details.

pydantic_evals/dataset.py CHANGED
@@ -50,16 +50,6 @@ if sys.version_info < (3, 11):
50
50
  else:
51
51
  ExceptionGroup = ExceptionGroup # pragma: lax no cover
52
52
 
53
- # while waiting for https://github.com/pydantic/logfire/issues/745
54
- try:
55
- import logfire._internal.stack_info
56
- except ImportError:
57
- pass
58
- else:
59
- from pathlib import Path
60
-
61
- logfire._internal.stack_info.NON_USER_CODE_PREFIXES += (str(Path(__file__).parent.absolute()),) # pyright: ignore[reportPrivateImportUsage]
62
-
63
53
  __all__ = (
64
54
  'Case',
65
55
  'Dataset',
@@ -859,11 +849,10 @@ async def _run_task(
859
849
  _CURRENT_TASK_RUN.reset(token)
860
850
 
861
851
  if isinstance(span_tree, SpanTree): # pragma: no branch
862
- # TODO: Question: Should we make this metric-attributes functionality more user-configurable in some way before merging?
863
- # Note: the use of otel for collecting these metrics is the main reason why I think we should require at least otel as a dependency, if not logfire;
864
- # otherwise, we don't have a great way to get usage data from arbitrary frameworks.
865
- # Ideally we wouldn't need to hard-code the specific logic here, but I'm not sure a great way to expose it to
866
- # users. Maybe via an argument of type Callable[[SpanTree], dict[str, int | float]] or similar?
852
+ # Idea for making this more configurable: replace the following logic with a call to a user-provided function
853
+ # of type Callable[[_TaskRun, SpanTree], None] or similar, (maybe no _TaskRun and just use the public APIs).
854
+ # That way users can customize this logic. We'd default to a function that does the current thing but also
855
+ # allow `None` to disable it entirely.
867
856
  for node in span_tree:
868
857
  if node.attributes.get('gen_ai.operation.name') == 'chat':
869
858
  task_run.increment_metric('requests', 1)
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: pydantic-evals
3
- Version: 0.8.0
3
+ Version: 0.8.1
4
4
  Summary: Framework for evaluating stochastic code execution, especially code making use of LLMs
5
5
  Project-URL: Homepage, https://ai.pydantic.dev/evals
6
6
  Project-URL: Source, https://github.com/pydantic/pydantic-ai
@@ -32,7 +32,7 @@ Requires-Python: >=3.9
32
32
  Requires-Dist: anyio>=0
33
33
  Requires-Dist: eval-type-backport>=0; python_version < '3.11'
34
34
  Requires-Dist: logfire-api>=3.14.1
35
- Requires-Dist: pydantic-ai-slim==0.8.0
35
+ Requires-Dist: pydantic-ai-slim==0.8.1
36
36
  Requires-Dist: pydantic>=2.10
37
37
  Requires-Dist: pyyaml>=6.0.2
38
38
  Requires-Dist: rich>=13.9.4
@@ -1,6 +1,6 @@
1
1
  pydantic_evals/__init__.py,sha256=OKRbfhdc8UZPzrPJMZUQwvzIxLhXmEZxz1ZuD921fy4,839
2
2
  pydantic_evals/_utils.py,sha256=PfhmPbdQp-q90s568LuG45zDDXxgO13BEz8MQJK8qw4,2922
3
- pydantic_evals/dataset.py,sha256=NSEKJLqKfVVu0YpmPJsb6rygwgglyc7h6_dIZAsKqqk,47036
3
+ pydantic_evals/dataset.py,sha256=tyQ59bV7V_1R1f0pdIbTcbdAYkKF6-Ype6EbdKrP2Lg,46508
4
4
  pydantic_evals/generation.py,sha256=Yd1rfbsDjjBBHDk-1KDu48hlITjM2-74rTnPBD_sqbA,3494
5
5
  pydantic_evals/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
6
6
  pydantic_evals/evaluators/__init__.py,sha256=k_sTzRezIOUj7L2x3WC0_r8943jRo72uqb2vKplD5EU,660
@@ -17,7 +17,7 @@ pydantic_evals/otel/_errors.py,sha256=aW1414eTofpA7R_DUgOeT-gj7YA6OXmm8Y4oYeFukD
17
17
  pydantic_evals/otel/span_tree.py,sha256=LV5Hsyo4riJzevHyBz8wxP82S-ry5zeKYi9bKWjGCS8,23057
18
18
  pydantic_evals/reporting/__init__.py,sha256=J6xdy6b1MG1kk47Za3oKu0c-RXD8jEi2TYGkRU1_qZU,43396
19
19
  pydantic_evals/reporting/render_numbers.py,sha256=8SKlK3etbD7HnSWWHCE993ceCNLZCepVQ-SsqUIhyxk,6916
20
- pydantic_evals-0.8.0.dist-info/METADATA,sha256=p_7oTxKt9CbIQfbyt0m9HPLNzYo8qRNm5NGid_pSOhg,7942
21
- pydantic_evals-0.8.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
22
- pydantic_evals-0.8.0.dist-info/licenses/LICENSE,sha256=vA6Jc482lEyBBuGUfD1pYx-cM7jxvLYOxPidZ30t_PQ,1100
23
- pydantic_evals-0.8.0.dist-info/RECORD,,
20
+ pydantic_evals-0.8.1.dist-info/METADATA,sha256=hgMkoJYEE-7S700P6Q-GSEGC0Xar0QA8f4ZRLyRCsfU,7942
21
+ pydantic_evals-0.8.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
22
+ pydantic_evals-0.8.1.dist-info/licenses/LICENSE,sha256=vA6Jc482lEyBBuGUfD1pYx-cM7jxvLYOxPidZ30t_PQ,1100
23
+ pydantic_evals-0.8.1.dist-info/RECORD,,