mteb 2.7.16__py3-none-any.whl → 2.7.18__py3-none-any.whl

mteb/tasks/retrieval/multilingual/cur_ev1_retrieval.py

@@ -111,7 +111,7 @@ class CUREv1Retrieval(AbsTaskRetrieval):
 
         return queries
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/jina_vdr_bench_retrieval.py

@@ -148,7 +148,7 @@ def _load_data(
     return corpus, queries, relevant_docs
 
 
-def load_data(self, num_proc: int = 1, **kwargs) -> None:
+def load_data(self, num_proc: int | None = None, **kwargs) -> None:
     if self.data_loaded:
         return
 

mteb/tasks/retrieval/multilingual/miracl_vision_retrieval.py

@@ -143,7 +143,7 @@ class MIRACLVisionRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that is relevant to the user's query."},
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/mr_tidy_retrieval.py

@@ -108,7 +108,7 @@ class MrTidyRetrieval(AbsTaskRetrieval):
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/public_health_qa_retrieval.py

@@ -97,7 +97,7 @@ class PublicHealthQARetrieval(AbsTaskRetrieval):
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/ru_sci_bench_retrieval.py

@@ -103,7 +103,7 @@ class RuSciBenchCiteRetrieval(AbsTaskRetrieval):
         },
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 
@@ -161,7 +161,7 @@ class RuSciBenchCociteRetrieval(AbsTaskRetrieval):
         },
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/statcan_dialogue_dataset_retrieval.py

@@ -96,7 +96,7 @@ de Vries, Harm},
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/vdr_multilingual_retrieval.py

@@ -126,7 +126,7 @@ class VDRMultilingualRetrieval(AbsTaskRetrieval):
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/vidore2_bench_retrieval.py

@@ -16,7 +16,7 @@ def _load_data(
     splits: list[str],
     langs: list | None = None,
     revision: str | None = None,
-    num_proc: int = 1,
+    num_proc: int | None = None,
 ):
     if langs is None:
         corpus = {}
@@ -131,7 +131,7 @@ class Vidore2ESGReportsRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that relevant to the user's question."},
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 
@@ -179,7 +179,7 @@ class Vidore2EconomicsReportsRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that relevant to the user's question."},
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 
@@ -227,7 +227,7 @@ class Vidore2BioMedicalLecturesRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that relevant to the user's question."},
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 
@@ -275,7 +275,7 @@ class Vidore2ESGReportsHLRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that relevant to the user's question."},
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/vidore3_bench_retrieval.py

@@ -68,6 +68,7 @@ class Vidore3FinanceFrRetrieval(AbsTaskRetrieval):
         license="cc-by-4.0",
         annotations_creators="derived",
         dialect=[],
+        modalities=["text", "image"],
         sample_creation="created and machine-translated",
         bibtex_citation=r"""
 @article{loison2026vidorev3comprehensiveevaluation,

mteb/tasks/retrieval/multilingual/wit_t2i_retrieval.py

@@ -116,7 +116,7 @@ class WITT2IRetrieval(AbsTaskRetrieval):
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/x_flickr30k_co_t2i_retrieval.py

@@ -104,7 +104,7 @@ class XFlickr30kCoT2IRetrieval(AbsTaskRetrieval):
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/x_qu_ad_retrieval.py

@@ -64,7 +64,7 @@ class XQuADRetrieval(AbsTaskRetrieval):
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/multilingual/xm3600_t2i_retrieval.py

@@ -146,7 +146,7 @@ class XM3600T2IRetrieval(AbsTaskRetrieval):
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_android_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackAndroidNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackAndroid"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_english_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackEnglishNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackEnglish"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_gaming_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackGamingNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackGamingRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_gis_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackGisNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackGisRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_mathematica_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackMathematicaNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackMathematicaRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_physics_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackPhysicsNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackPhysicsRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_programmers_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackProgrammersNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackProgrammersRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_stats_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackStatsNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackStatsRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_tex_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackTexNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackTexRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_unix_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackUnixNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackUnixRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_webmasters_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackWebmastersNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackWebmastersRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nld/cqa_dupstack_wordpress_nl_retrieval.py

@@ -42,7 +42,7 @@ class CQADupstackWordpressNLRetrieval(AbsTaskRetrieval):
         adapted_from=["CQADupstackWordpressRetrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/retrieval/nob/norquad.py

@@ -50,7 +50,7 @@ Fishel, Mark},
         },
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         """Load dataset from HuggingFace hub"""
         if self.data_loaded:
             return
@@ -58,7 +58,7 @@ Fishel, Mark},
         self.dataset_transform()
         self.data_loaded = True
 
-    def dataset_transform(self, num_proc: int = 1, **kwargs) -> None:
+    def dataset_transform(self, num_proc: int | None = None, **kwargs) -> None:
         """And transform to a retrieval dataset, which have the following attributes
 
         self.corpus = dict[doc_id, dict[str, str]] #id => dict with document data like title and text

mteb/tasks/retrieval/nob/snl_retrieval.py

@@ -37,7 +37,7 @@ class SNLRetrieval(AbsTaskRetrieval):
         task_subtypes=["Article retrieval"],
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         """Load dataset from HuggingFace hub"""
         if self.data_loaded:
             return
@@ -45,7 +45,7 @@ class SNLRetrieval(AbsTaskRetrieval):
         self.dataset_transform()
         self.data_loaded = True
 
-    def dataset_transform(self, num_proc: int = 1, **kwargs) -> None:
+    def dataset_transform(self, num_proc: int | None = None, **kwargs) -> None:
         """And transform to a retrieval dataset, which have the following attributes
 
         self.corpus = dict[doc_id, dict[str, str]] #id => dict with document data like title and text

mteb/tasks/retrieval/slk/slovak_sum_retrieval.py

@@ -36,7 +36,7 @@ class SlovakSumRetrieval(AbsTaskRetrieval):
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
         self.corpus, self.queries, self.relevant_docs = {}, {}, {}

mteb/tasks/retrieval/vie/vie_qu_ad_retrieval.py

@@ -52,7 +52,7 @@ Zong, Chengqing},
 """,
     )
 
-    def load_data(self, num_proc: int = 1, **kwargs) -> None:
+    def load_data(self, num_proc: int | None = None, **kwargs) -> None:
         if self.data_loaded:
             return
 

mteb/tasks/sts/multilingual/sem_rel24_sts.py

@@ -66,6 +66,6 @@ Seid Muhie Yimam and Saif M. Mohammad},
     min_score = 0
     max_score = 1
 
-    def dataset_transform(self, num_proc: int = 1, **kwargs) -> None:
+    def dataset_transform(self, num_proc: int | None = None, **kwargs) -> None:
         for lang, subset in self.dataset.items():
             self.dataset[lang] = subset.rename_column("label", "score")

mteb/tasks/sts/multilingual/sts_benchmark_multilingual_sts.py

@@ -56,6 +56,6 @@ class STSBenchmarkMultilingualSTS(AbsTaskSTS):
     min_score = 0
     max_score = 5
 
-    def dataset_transform(self, num_proc: int = 1, **kwargs) -> None:
+    def dataset_transform(self, num_proc: int | None = None, **kwargs) -> None:
         for lang, subset in self.dataset.items():
             self.dataset[lang] = subset.rename_column("similarity_score", "score")

mteb/tasks/sts/por/assin2_sts.py

@@ -39,7 +39,7 @@ class Assin2STS(AbsTaskSTS):
     min_score = 1
     max_score = 5
 
-    def dataset_transform(self, num_proc: int = 1, **kwargs) -> None:
+    def dataset_transform(self, num_proc: int | None = None, **kwargs) -> None:
         self.dataset = self.dataset.rename_columns(
             {
                 "premise": "sentence1",

mteb/types/_encoder_io.py

@@ -7,6 +7,7 @@ from typing import TYPE_CHECKING, TypedDict
 import numpy as np
 import torch
 from datasets import Dataset
+from numpy.typing import NDArray
 
 if TYPE_CHECKING:
     from PIL import Image
@@ -26,8 +27,8 @@ class EncodeKwargs(TypedDict):
 
 
 # --- Output types ---
-Array = np.ndarray | torch.Tensor
-"""General array type, can be a numpy array or a torch tensor."""
+Array = NDArray[np.floating | np.integer | np.bool] | torch.Tensor
+"""General array type, can be a numpy array (float, int, or bool) or a torch tensor."""
 
 
 # --- Input types ---

{mteb-2.7.16.dist-info → mteb-2.7.18.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mteb
-Version: 2.7.16
+Version: 2.7.18
 Summary: Massive Text Embedding Benchmark
 Author-email: MTEB Contributors <niklas@huggingface.co>, Kenneth Enevoldsen <kenneth.enevoldsen@cas.au.dk>, Nouamane Tazi <nouamane@huggingface.co>, Nils Reimers <info@nils-reimers.de>
 Maintainer-email: Kenneth Enevoldsen <kenneth.enevoldsen@cas.au.dk>, Roman Solomatin <risolomatin@gmail.com>, Isaac Chung <chungisaac1217@gmail.com>