evalproof 0.2.0__tar.gz → 0.2.1__tar.gz
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.
- {evalproof-0.2.0 → evalproof-0.2.1}/CHANGELOG.md +12 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/CITATION.cff +1 -1
- {evalproof-0.2.0 → evalproof-0.2.1}/PKG-INFO +1 -3
- {evalproof-0.2.0 → evalproof-0.2.1}/README.md +2 -2
- {evalproof-0.2.0 → evalproof-0.2.1}/pyproject.toml +6 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/_version.py +1 -1
- {evalproof-0.2.0 → evalproof-0.2.1}/.gitignore +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/LICENSE +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/NOTICE +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/__init__.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/cache.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/calibration.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/cli.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/compare.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/config.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/dataset.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/gate.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/models.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/providers/__init__.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/providers/anthropic.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/providers/base.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/providers/ollama.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/providers/openai.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/py.typed +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/report.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/runner.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/scorers/__init__.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/scorers/deterministic.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/scorers/judge.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/scorers/retrieval.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/stats.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/src/evalproof/systems.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/__init__.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/conftest.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_cache_systems.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_calibration.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_compare_gate_report.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_config_dataset.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_providers.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_public_api.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_runner_cli.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_scorers.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_security.py +0 -0
- {evalproof-0.2.0 → evalproof-0.2.1}/tests/test_stats.py +0 -0
|
@@ -2,6 +2,17 @@
|
|
|
2
2
|
|
|
3
3
|
All notable changes are listed here. The project follows [Semantic Versioning](https://semver.org/).
|
|
4
4
|
|
|
5
|
+
## [0.2.1] — 2026-10-16
|
|
6
|
+
|
|
7
|
+
### Fixed
|
|
8
|
+
|
|
9
|
+
- The PyPI and Python badges on the 0.2.0 PyPI page read "package or version not found".
|
|
10
|
+
PyPI's image proxy fetched them about a minute after the first upload, before the badge
|
|
11
|
+
service could see the package, and keeps that answer for 12 to 24 hours. The PyPI page
|
|
12
|
+
now leaves those two badges out, since it shows the version and supported Pythons
|
|
13
|
+
itself, and the README's copies use new URLs so no cache serves the old images. CI fails
|
|
14
|
+
if a PyPI-data badge reaches the PyPI description again.
|
|
15
|
+
|
|
5
16
|
## [0.2.0] — 2026-10-16
|
|
6
17
|
|
|
7
18
|
The first release on PyPI: `pip install evalproof`.
|
|
@@ -93,6 +104,7 @@ First release.
|
|
|
93
104
|
- Two examples: an offline ticket router (CI's gate contract) and a support assistant for
|
|
94
105
|
a fictional product judged by a local model, with measurements.
|
|
95
106
|
|
|
107
|
+
[0.2.1]: https://github.com/technobasant/evalproof/compare/v0.2.0...v0.2.1
|
|
96
108
|
[0.2.0]: https://github.com/technobasant/evalproof/compare/v0.1.2...v0.2.0
|
|
97
109
|
[0.1.2]: https://github.com/technobasant/evalproof/compare/v0.1.1...v0.1.2
|
|
98
110
|
[0.1.1]: https://github.com/technobasant/evalproof/compare/v0.1.0...v0.1.1
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.5
|
|
2
2
|
Name: evalproof
|
|
3
|
-
Version: 0.2.
|
|
3
|
+
Version: 0.2.1
|
|
4
4
|
Summary: Treat LLM evaluation as measurement: calibrate the judge, correct the pass rate, gate CI on significant regressions.
|
|
5
5
|
Project-URL: Homepage, https://github.com/technobasant/evalproof
|
|
6
6
|
Project-URL: Documentation, https://github.com/technobasant/evalproof#readme
|
|
@@ -38,8 +38,6 @@ Description-Content-Type: text/markdown
|
|
|
38
38
|
|
|
39
39
|
# evalproof
|
|
40
40
|
|
|
41
|
-
[](https://pypi.org/project/evalproof/)
|
|
42
|
-
[](https://pypi.org/project/evalproof/)
|
|
43
41
|
[](https://github.com/technobasant/evalproof/actions/workflows/ci.yml)
|
|
44
42
|
[](https://github.com/technobasant/evalproof/blob/main/LICENSE)
|
|
45
43
|
[](https://evalproof.streamlit.app)
|
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
# evalproof
|
|
2
2
|
|
|
3
|
-
[](https://pypi.org/project/evalproof/)
|
|
4
|
-
[](https://pypi.org/project/evalproof/)
|
|
3
|
+
[](https://pypi.org/project/evalproof/)
|
|
4
|
+
[](https://pypi.org/project/evalproof/)
|
|
5
5
|
[](https://github.com/technobasant/evalproof/actions/workflows/ci.yml)
|
|
6
6
|
[](LICENSE)
|
|
7
7
|
[](https://evalproof.streamlit.app)
|
|
@@ -67,6 +67,12 @@ content-type = "text/markdown"
|
|
|
67
67
|
[[tool.hatch.metadata.hooks.fancy-pypi-readme.fragments]]
|
|
68
68
|
path = "README.md"
|
|
69
69
|
|
|
70
|
+
# The PyPI page shows its own version and supported Pythons, and a badge rendered there
|
|
71
|
+
# from a just-uploaded release can be cached as "not found" for a day.
|
|
72
|
+
[[tool.hatch.metadata.hooks.fancy-pypi-readme.substitutions]]
|
|
73
|
+
pattern = '\[!\[(PyPI|Python)\]\(https://img\.shields\.io/pypi/[^)]+\)\]\([^)]+\)\n'
|
|
74
|
+
replacement = ''
|
|
75
|
+
|
|
70
76
|
[[tool.hatch.metadata.hooks.fancy-pypi-readme.substitutions]]
|
|
71
77
|
pattern = '(?s)```mermaid\n.*?```'
|
|
72
78
|
replacement = "The architecture diagram is in the [README on GitHub](https://github.com/technobasant/evalproof#architecture)."
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|