mteb 2.7.3__py3-none-any.whl → 2.7.5__py3-none-any.whl

mteb/tasks/retrieval/multilingual/vidore2_bench_retrieval.py

@@ -16,6 +16,7 @@ def _load_data(
     splits: list[str],
     langs: list | None = None,
     revision: str | None = None,
+    num_proc: int = 1,
 ):
     if langs is None:
         corpus = {}
@@ -32,6 +33,7 @@ def _load_data(
             "queries",
             split=split,
             revision=revision,
+            num_proc=num_proc,
         )
         query_ds = query_ds.map(
             lambda x: {
@@ -40,6 +42,7 @@ def _load_data(
                 "modality": "text",
             },
             remove_columns=["query-id", "query"],
+            num_proc=num_proc,
         )
 
         corpus_ds = load_dataset(
@@ -47,6 +50,7 @@ def _load_data(
             "corpus",
             split=split,
             revision=revision,
+            num_proc=num_proc,
         )
         corpus_ds = corpus_ds.map(
             lambda x: {
@@ -54,6 +58,7 @@ def _load_data(
                 "modality": "image",
             },
             remove_columns=["corpus-id"],
+            num_proc=num_proc,
         )
         corpus_ds = corpus_ds.select_columns(["id", "image"])
 
@@ -62,6 +67,7 @@ def _load_data(
             "qrels",
             split=split,
             revision=revision,
+            num_proc=num_proc,
         )
 
         if langs is None:
@@ -125,7 +131,7 @@ class Vidore2ESGReportsRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that relevant to the user's question."},
     )
 
-    def load_data(self) -> None:
+    def load_data(self, num_proc: int = 1, **kwargs) -> None:
         if self.data_loaded:
             return
 
@@ -134,6 +140,7 @@ class Vidore2ESGReportsRetrieval(AbsTaskRetrieval):
             splits=self.metadata.eval_splits,
             langs=_LANGS.keys(),
             revision=self.metadata.dataset["revision"],
+            num_proc=num_proc,
         )
 
         self.data_loaded = True
@@ -172,7 +179,7 @@ class Vidore2EconomicsReportsRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that relevant to the user's question."},
     )
 
-    def load_data(self) -> None:
+    def load_data(self, num_proc: int = 1, **kwargs) -> None:
         if self.data_loaded:
             return
 
@@ -181,6 +188,7 @@ class Vidore2EconomicsReportsRetrieval(AbsTaskRetrieval):
             splits=self.metadata.eval_splits,
             langs=_LANGS.keys(),
             revision=self.metadata.dataset["revision"],
+            num_proc=num_proc,
         )
 
         self.data_loaded = True
@@ -219,7 +227,7 @@ class Vidore2BioMedicalLecturesRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that relevant to the user's question."},
     )
 
-    def load_data(self) -> None:
+    def load_data(self, num_proc: int = 1, **kwargs) -> None:
         if self.data_loaded:
             return
 
@@ -228,6 +236,7 @@ class Vidore2BioMedicalLecturesRetrieval(AbsTaskRetrieval):
             splits=self.metadata.eval_splits,
             langs=_LANGS.keys(),
             revision=self.metadata.dataset["revision"],
+            num_proc=num_proc,
         )
 
         self.data_loaded = True
@@ -266,7 +275,7 @@ class Vidore2ESGReportsHLRetrieval(AbsTaskRetrieval):
         prompt={"query": "Find a screenshot that relevant to the user's question."},
     )
 
-    def load_data(self) -> None:
+    def load_data(self, num_proc: int = 1, **kwargs) -> None:
         if self.data_loaded:
             return
 
@@ -274,6 +283,7 @@ class Vidore2ESGReportsHLRetrieval(AbsTaskRetrieval):
             path=self.metadata.dataset["path"],
             splits=self.metadata.eval_splits,
             revision=self.metadata.dataset["revision"],
+            num_proc=num_proc,
         )
 
         self.data_loaded = True

mteb/tasks/retrieval/multilingual/wit_t2i_retrieval.py

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

mteb/tasks/retrieval/nob/norquad.py

@@ -50,7 +50,7 @@ Fishel, Mark},
         },
     )
 
-    def load_data(self) -> None:
+    def load_data(self, num_proc: int = 1, **kwargs) -> None:
         """Load dataset from HuggingFace hub"""
         if self.data_loaded:
             return

mteb/tasks/retrieval/nob/snl_retrieval.py

@@ -37,7 +37,7 @@ class SNLRetrieval(AbsTaskRetrieval):
         task_subtypes=["Article retrieval"],
     )
 
-    def load_data(self) -> None:
+    def load_data(self, num_proc: int = 1, **kwargs) -> None:
         """Load dataset from HuggingFace hub"""
         if self.data_loaded:
             return

mteb/tasks/retrieval/slk/slovak_sum_retrieval.py

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

mteb/tasks/sts/fao/faroese_sts.py

@@ -43,5 +43,5 @@ Vulić, Ivan},
     min_score = 0
     max_score = 5
 
-    def dataset_transform(self):
+    def dataset_transform(self, num_proc: int = 1):
         self.dataset = self.dataset.rename_column("label", "score")

mteb/tasks/sts/fra/sick_fr_sts.py

@@ -30,7 +30,7 @@ class SickFrSTS(AbsTaskSTS):
     min_score = 0
     max_score = 5
 
-    def dataset_transform(self):
+    def dataset_transform(self, num_proc: int = 1):
         self.dataset = self.dataset.rename_columns(
             {
                 "sentence_A": "sentence1",

mteb/tasks/sts/kor/klue_sts.py

@@ -40,7 +40,7 @@ class KlueSTS(AbsTaskSTS):
     min_score = 0
     max_score = 5
 
-    def dataset_transform(self):
+    def dataset_transform(self, num_proc: int = 1):
         # In the case of KLUE STS, score value is nested within the `labels` field.
         # We need to extract the `score` and move it outside of the `labels` field for access.
         for split in self.dataset:

mteb/tasks/sts/por/sick_br_sts.py

@@ -52,7 +52,7 @@ and de Paiva, Valeria},
     min_score = 1
     max_score = 5
 
-    def dataset_transform(self):
+    def dataset_transform(self, num_proc: int = 1):
         self.dataset = self.stratified_subsampling(
             self.dataset,
             seed=42,

mteb/tasks/sts/rus/ru_para_phraser_sts.py

@@ -54,7 +54,7 @@ Filippskikh, Elizaveta},
     min_score = -1
     max_score = 1
 
-    def dataset_transform(self):
+    def dataset_transform(self, num_proc: int = 1):
         self.dataset = self.dataset.rename_columns(
             {
                 "text_1": "sentence1",

mteb/tasks/zeroshot_classification/eng/sci_mmir.py

@@ -41,7 +41,7 @@ class SciMMIR(AbsTaskZeroShotClassification):
 
     label_column_name: str = "class"
 
-    def dataset_transform(self):
+    def dataset_transform(self, num_proc: int = 1):
         class_code = {
             "fig_result": 0,
             "fig_illustration": 1,

{mteb-2.7.3.dist-info → mteb-2.7.5.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mteb
-Version: 2.7.3
+Version: 2.7.5
 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>