valor-lite 0.34.0__tar.gz → 0.34.2__tar.gz

{valor_lite-0.34.0 → valor_lite-0.34.2}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: valor-lite
-Version: 0.34.0
+Version: 0.34.2
 Summary: Evaluate machine learning models.
 Project-URL: homepage, https://www.striveworks.com
 Requires-Python: >=3.10
@@ -8,17 +8,14 @@ Description-Content-Type: text/markdown
 Requires-Dist: numpy
 Requires-Dist: tqdm
 Requires-Dist: shapely
-Requires-Dist: evaluate
-Requires-Dist: nltk
-Requires-Dist: rouge_score
+Provides-Extra: nlp
+Requires-Dist: evaluate; extra == "nlp"
+Requires-Dist: nltk; extra == "nlp"
+Requires-Dist: rouge_score; extra == "nlp"
 Provides-Extra: mistral
 Requires-Dist: mistralai>=1.0; extra == "mistral"
 Provides-Extra: openai
 Requires-Dist: openai; extra == "openai"
-Provides-Extra: test
-Requires-Dist: pytest; extra == "test"
-Requires-Dist: coverage; extra == "test"
-Requires-Dist: pre-commit; extra == "test"
 Provides-Extra: docs
 Requires-Dist: mkdocs; extra == "docs"
 Requires-Dist: mkdocs-material; extra == "docs"
@@ -26,8 +23,14 @@ Requires-Dist: mkdocstrings; extra == "docs"
 Requires-Dist: mkdocstrings-python; extra == "docs"
 Requires-Dist: mkdocs-include-dir-to-nav; extra == "docs"
 Requires-Dist: mkdocs-swagger-ui-tag; extra == "docs"
+Provides-Extra: test
+Requires-Dist: pytest; extra == "test"
+Requires-Dist: coverage; extra == "test"
+Requires-Dist: pre-commit; extra == "test"
+Provides-Extra: benchmark
+Requires-Dist: requests; extra == "benchmark"
 Provides-Extra: dev
-Requires-Dist: valor-lite[docs,mistral,openai,test]; extra == "dev"
+Requires-Dist: valor-lite[benchmark,docs,mistral,nlp,openai,test]; extra == "dev"
 
 # valor-lite: Fast, local machine learning evaluation.
 

{valor_lite-0.34.0 → valor_lite-0.34.2}/pyproject.toml

@@ -9,9 +9,6 @@ dependencies = [
     "numpy",
     "tqdm",
     "shapely",
-    "evaluate",
-    "nltk",
-    "rouge_score",
 ]
 
 [project.urls]
@@ -22,9 +19,9 @@ requires = ["setuptools>=61.0", "setuptools_scm[toml]>=6.2"]
 build-backend = "setuptools.build_meta"
 
 [project.optional-dependencies]
+nlp = ["evaluate", "nltk", "rouge_score"]
 mistral = ["mistralai >= 1.0"]
 openai = ["openai"]
-test = ["pytest", "coverage", "pre-commit"]
 docs = [
     "mkdocs",
     "mkdocs-material",
@@ -33,7 +30,9 @@ docs = [
     "mkdocs-include-dir-to-nav",
     "mkdocs-swagger-ui-tag",
 ]
-dev = ["valor-lite[openai, mistral, test, docs]"]
+test = ["pytest", "coverage", "pre-commit"]
+benchmark = ["requests"]
+dev = ["valor-lite[nlp, openai, mistral, benchmark, test, docs]"]
 
 [tool.black]
 line-length = 79

{valor_lite-0.34.0 → valor_lite-0.34.2}/valor_lite/classification/computation.py

@@ -1,6 +1,8 @@
 import numpy as np
 from numpy.typing import NDArray
 
+import valor_lite.classification.numpy_compatibility as npc
+
 
 def _compute_rocauc(
     data: NDArray[np.float64],
@@ -9,7 +11,7 @@ def _compute_rocauc(
     n_labels: int,
     mask_matching_labels: NDArray[np.bool_],
     pd_labels: NDArray[np.int32],
-):
+) -> tuple[NDArray[np.float64], float]:
     """
     Compute ROCAUC and mean ROCAUC.
     """
@@ -56,12 +58,12 @@ def _compute_rocauc(
     np.maximum.accumulate(tpr, axis=1, out=tpr)
 
     # compute rocauc
-    rocauc = np.trapz(x=fpr, y=tpr, axis=1)  # type: ignore - numpy will be switching to `trapezoid` in the future.
+    rocauc = npc.trapezoid(x=fpr, y=tpr, axis=1)
 
     # compute mean rocauc
     mean_rocauc = rocauc.mean()
 
-    return rocauc, mean_rocauc
+    return rocauc, mean_rocauc  # type: ignore[reportReturnType]
 
 
 def compute_precision_recall_rocauc(

valor_lite-0.34.2/valor_lite/classification/numpy_compatibility.py

@@ -0,0 +1,13 @@
+import numpy as np
+from numpy.typing import NDArray
+
+try:
+    _numpy_trapezoid = np.trapezoid  # numpy v2
+except AttributeError:
+    _numpy_trapezoid = np.trapz  # numpy v1
+
+
+def trapezoid(
+    x: NDArray[np.float64], y: NDArray[np.float64], axis: int
+) -> NDArray[np.float64]:
+    return _numpy_trapezoid(x=x, y=y, axis=axis)  # type: ignore - NumPy compatibility

{valor_lite-0.34.0 → valor_lite-0.34.2}/valor_lite/object_detection/computation.py

@@ -560,12 +560,12 @@ def compute_precion_recall(
     )
 
     return (
-        ap_results,
+        ap_results,  # type: ignore[reportReturnType]
         ar_results,
         accuracy,
         counts,
         pr_curve,
-    )  # type: ignore[reportReturnType]
+    )
 
 
 def _count_with_examples(

{valor_lite-0.34.0 → valor_lite-0.34.2}/valor_lite/object_detection/manager.py

@@ -195,8 +195,8 @@ class Evaluator:
 
     def compute_precision_recall(
         self,
-        iou_thresholds: list[float] = [0.5, 0.75, 0.9],
-        score_thresholds: list[float] = [0.5],
+        iou_thresholds: list[float],
+        score_thresholds: list[float],
         filter_: Filter | None = None,
     ) -> dict[MetricType, list[Metric]]:
         """
@@ -240,9 +240,9 @@ class Evaluator:
 
     def compute_confusion_matrix(
         self,
-        iou_thresholds: list[float] = [0.5, 0.75, 0.9],
-        score_thresholds: list[float] = [0.5],
-        number_of_examples: int = 0,
+        iou_thresholds: list[float],
+        score_thresholds: list[float],
+        number_of_examples: int,
         filter_: Filter | None = None,
     ) -> list[Metric]:
         """
@@ -254,7 +254,7 @@ class Evaluator:
             A list of IOU thresholds to compute metrics over.
         score_thresholds : list[float]
             A list of score thresholds to compute metrics over.
-        number_of_examples : int, default=0
+        number_of_examples : int
             Maximum number of annotation examples to return in ConfusionMatrix.
         filter_ : Filter, optional
             An optional filter object.
@@ -295,7 +295,7 @@ class Evaluator:
 
     def evaluate(
         self,
-        iou_thresholds: list[float] = [0.5, 0.75, 0.9],
+        iou_thresholds: list[float] = [0.1, 0.5, 0.75],
         score_thresholds: list[float] = [0.5],
         number_of_examples: int = 0,
         filter_: Filter | None = None,
@@ -305,9 +305,9 @@ class Evaluator:
 
         Parameters
         ----------
-        iou_thresholds : list[float]
+        iou_thresholds : list[float], default=[0.1, 0.5, 0.75]
             A list of IOU thresholds to compute metrics over.
-        score_thresholds : list[float]
+        score_thresholds : list[float], default=[0.5]
             A list of score thresholds to compute metrics over.
         number_of_examples : int, default=0
             Maximum number of annotation examples to return in ConfusionMatrix.

{valor_lite-0.34.0 → valor_lite-0.34.2}/valor_lite/text_generation/computation.py

@@ -1,7 +1,3 @@
-import evaluate
-from nltk.tokenize import RegexpTokenizer
-from nltk.translate import bleu_score
-
 from valor_lite.text_generation.llm.generation import (
     generate_answer_correctness_verdicts,
     generate_answer_relevance_verdicts,
@@ -551,6 +547,8 @@ def calculate_rouge_scores(
     use_stemmer: bool, default=False
         If True, uses Porter stemmer to strip word suffixes. Defaults to False.
     """
+    import evaluate
+
     rouge = evaluate.load("rouge")
 
     metrics = rouge.compute(
@@ -589,6 +587,9 @@ def calculate_sentence_bleu(
         higher/lower order ngrams, use customized weights. Example: when accounting
         for up to 5-grams with uniform weights (this is called BLEU-5) use [1/5]*5
     """
+    from nltk.tokenize import RegexpTokenizer
+    from nltk.translate import bleu_score
+
     if len(weights) == 0:
         raise ValueError("At least one weight should be defined.")
 

{valor_lite-0.34.0 → valor_lite-0.34.2}/valor_lite.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: valor-lite
-Version: 0.34.0
+Version: 0.34.2
 Summary: Evaluate machine learning models.
 Project-URL: homepage, https://www.striveworks.com
 Requires-Python: >=3.10
@@ -8,17 +8,14 @@ Description-Content-Type: text/markdown
 Requires-Dist: numpy
 Requires-Dist: tqdm
 Requires-Dist: shapely
-Requires-Dist: evaluate
-Requires-Dist: nltk
-Requires-Dist: rouge_score
+Provides-Extra: nlp
+Requires-Dist: evaluate; extra == "nlp"
+Requires-Dist: nltk; extra == "nlp"
+Requires-Dist: rouge_score; extra == "nlp"
 Provides-Extra: mistral
 Requires-Dist: mistralai>=1.0; extra == "mistral"
 Provides-Extra: openai
 Requires-Dist: openai; extra == "openai"
-Provides-Extra: test
-Requires-Dist: pytest; extra == "test"
-Requires-Dist: coverage; extra == "test"
-Requires-Dist: pre-commit; extra == "test"
 Provides-Extra: docs
 Requires-Dist: mkdocs; extra == "docs"
 Requires-Dist: mkdocs-material; extra == "docs"
@@ -26,8 +23,14 @@ Requires-Dist: mkdocstrings; extra == "docs"
 Requires-Dist: mkdocstrings-python; extra == "docs"
 Requires-Dist: mkdocs-include-dir-to-nav; extra == "docs"
 Requires-Dist: mkdocs-swagger-ui-tag; extra == "docs"
+Provides-Extra: test
+Requires-Dist: pytest; extra == "test"
+Requires-Dist: coverage; extra == "test"
+Requires-Dist: pre-commit; extra == "test"
+Provides-Extra: benchmark
+Requires-Dist: requests; extra == "benchmark"
 Provides-Extra: dev
-Requires-Dist: valor-lite[docs,mistral,openai,test]; extra == "dev"
+Requires-Dist: valor-lite[benchmark,docs,mistral,nlp,openai,test]; extra == "dev"
 
 # valor-lite: Fast, local machine learning evaluation.
 

{valor_lite-0.34.0 → valor_lite-0.34.2}/valor_lite.egg-info/SOURCES.txt

@@ -14,6 +14,7 @@ valor_lite/classification/annotation.py
 valor_lite/classification/computation.py
 valor_lite/classification/manager.py
 valor_lite/classification/metric.py
+valor_lite/classification/numpy_compatibility.py
 valor_lite/classification/utilities.py
 valor_lite/object_detection/__init__.py
 valor_lite/object_detection/annotation.py

{valor_lite-0.34.0 → valor_lite-0.34.2}/valor_lite.egg-info/requires.txt

@@ -1,12 +1,12 @@
 numpy
 tqdm
 shapely
-evaluate
-nltk
-rouge_score
+
+[benchmark]
+requests
 
 [dev]
-valor-lite[docs,mistral,openai,test]
+valor-lite[benchmark,docs,mistral,nlp,openai,test]
 
 [docs]
 mkdocs
@@ -19,6 +19,11 @@ mkdocs-swagger-ui-tag
 [mistral]
 mistralai>=1.0
 
+[nlp]
+evaluate
+nltk
+rouge_score
+
 [openai]
 openai