mteb 2.7.15__py3-none-any.whl → 2.7.16__py3-none-any.whl

mteb/models/model_implementations/misc_models.py

@@ -1007,54 +1007,6 @@ thenlper__gte_small = ModelMeta(
   year={2023}
 }""",
 )
-OrlikB__KartonBERT_USE_base_v1 = ModelMeta(
-    name="OrlikB/KartonBERT-USE-base-v1",
-    model_type=["dense"],
-    revision="1f59dd58fe57995c0e867d5e29f03763eae99645",
-    release_date="2024-09-30",
-    languages=["pol-Latn"],
-    loader=sentence_transformers_loader,
-    n_parameters=103705344,
-    n_embedding_parameters=None,
-    memory_usage_mb=396,
-    max_tokens=512.0,
-    embed_dim=768,
-    license="gpl-3.0",
-    open_weights=True,
-    public_training_code=None,
-    public_training_data=None,
-    framework=["PyTorch"],
-    reference="https://huggingface.co/OrlikB/KartonBERT-USE-base-v1",
-    similarity_fn_name=ScoringFunction.COSINE,
-    use_instructions=None,
-    training_datasets=None,
-    adapted_from="KartonBERT-USE-base-v1",
-    superseded_by=None,
-)
-OrlikB__st_polish_kartonberta_base_alpha_v1 = ModelMeta(
-    name="OrlikB/st-polish-kartonberta-base-alpha-v1",
-    model_type=["dense"],
-    revision="5590a0e2d7bb43674e44d7076b3ff157f7d4a1cb",
-    release_date="2023-11-12",
-    languages=["pol-Latn"],
-    loader=sentence_transformers_loader,
-    n_parameters=None,
-    n_embedding_parameters=None,
-    memory_usage_mb=None,
-    max_tokens=514.0,
-    embed_dim=768,
-    license="lgpl",
-    open_weights=True,
-    public_training_code=None,
-    public_training_data=None,
-    framework=["PyTorch"],
-    reference="https://huggingface.co/OrlikB/st-polish-kartonberta-base-alpha-v1",
-    similarity_fn_name=ScoringFunction.COSINE,
-    use_instructions=None,
-    training_datasets=None,
-    adapted_from="st-polish-kartonberta-base-alpha-v1",
-    superseded_by=None,
-)
 sdadas__mmlw_e5_base = ModelMeta(
     name="sdadas/mmlw-e5-base",
     model_type=["dense"],

mteb/models/model_implementations/rerankers_custom.py

@@ -103,68 +103,6 @@ class BGEReranker(RerankerWrapper):
         return scores
 
 
-class MonoBERTReranker(RerankerWrapper):
-    name: str = "MonoBERT"
-
-    def __init__(
-        self,
-        model_name_or_path="castorini/monobert-large-msmarco",
-        torch_compile=False,
-        **kwargs,
-    ):
-        from transformers import AutoModelForSequenceClassification, AutoTokenizer
-
-        super().__init__(model_name_or_path, **kwargs)
-        if not self.device:
-            self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-        model_args = {}
-        if self.fp_options:
-            model_args["torch_dtype"] = self.fp_options
-        self.model = AutoModelForSequenceClassification.from_pretrained(
-            model_name_or_path,
-            **model_args,
-        )
-        self.model.to(self.device)
-        self.tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
-        self.max_length = self.tokenizer.model_max_length
-        logger.info(f"Using max_length of {self.max_length}")
-
-        self.model.eval()
-
-    @torch.inference_mode()
-    def predict(
-        self,
-        inputs1: DataLoader[BatchedInput],
-        inputs2: DataLoader[BatchedInput],
-        *,
-        task_metadata: TaskMetadata,
-        hf_split: str,
-        hf_subset: str,
-        prompt_type: PromptType | None = None,
-        **kwargs: Any,
-    ) -> Array:
-        queries = [text for batch in inputs1 for text in batch["query"]]
-        instructions = None
-        if "instruction" in inputs2.dataset.features:
-            instructions = [text for batch in inputs1 for text in batch["instruction"]]
-        passages = [text for batch in inputs2 for text in batch["text"]]
-
-        if instructions is not None and instructions[0] is not None:
-            queries = [f"{q} {i}".strip() for i, q in zip(instructions, queries)]
-
-        tokens = self.tokenizer(
-            queries,
-            passages,
-            padding=True,
-            truncation="only_second",
-            return_tensors="pt",
-            max_length=self.max_length,
-        ).to(self.device)
-        output = self.model(**tokens)[0]
-        batch_scores = torch.nn.functional.log_softmax(output, dim=1)
-        return batch_scores[:, 1].exp()
-
-
 class JinaReranker(RerankerWrapper):
     name = "Jina"
 
@@ -219,31 +157,6 @@ class JinaReranker(RerankerWrapper):
         return scores
 
 
-monobert_large = ModelMeta(
-    loader=MonoBERTReranker,
-    loader_kwargs=dict(
-        fp_options="float16",
-    ),
-    name="castorini/monobert-large-msmarco",
-    model_type=["cross-encoder"],
-    languages=["eng-Latn"],
-    open_weights=True,
-    revision="0a97706f3827389da43b83348d5d18c9d53876fa",
-    release_date="2020-05-28",
-    n_parameters=None,
-    n_embedding_parameters=31_254_528,
-    memory_usage_mb=None,
-    max_tokens=None,
-    embed_dim=None,
-    license=None,
-    public_training_code=None,
-    public_training_data=None,
-    similarity_fn_name=None,
-    use_instructions=None,
-    training_datasets=None,
-    framework=["Sentence Transformers", "PyTorch", "Transformers"],
-)
-
 # languages unclear: https://huggingface.co/jinaai/jina-reranker-v2-base-multilingual/discussions/28
 jina_reranker_multilingual = ModelMeta(
     loader=JinaReranker,

mteb/models/model_implementations/rerankers_monot5_based.py

@@ -34,7 +34,6 @@ prediction_tokens = {
     "unicamp-dl/mt5-base-en-msmarco": ["▁no", "▁yes"],
     "unicamp-dl/mt5-base-mmarco-v2": ["▁no", "▁yes"],
     "unicamp-dl/mt5-base-mmarco-v1": ["▁no", "▁yes"],
-    "unicamp-dl/mt5-13b-mmarco-100k": ["▁", "▁true"],
 }
 
 
@@ -919,28 +918,3 @@ mt5_base_mmarco_v2 = ModelMeta(
     use_instructions=None,
     framework=["PyTorch", "Transformers"],
 )
-
-mt5_13b_mmarco_100k = ModelMeta(
-    loader=MonoT5Reranker,
-    loader_kwargs=dict(
-        fp_options="float16",
-    ),
-    name="unicamp-dl/mt5-13b-mmarco-100k",
-    model_type=["cross-encoder"],
-    languages=mt5_languages,
-    open_weights=True,
-    revision="e1a4317e102a525ea9e16745ad21394a4f1bffbc",
-    release_date="2022-11-04",
-    n_parameters=None,
-    n_embedding_parameters=1_024_458_752,
-    memory_usage_mb=None,
-    max_tokens=None,
-    embed_dim=None,
-    license=None,
-    public_training_code=None,
-    public_training_data=None,
-    similarity_fn_name=None,
-    use_instructions=None,
-    training_datasets=None,
-    framework=["PyTorch", "Transformers"],
-)

{mteb-2.7.15.dist-info → mteb-2.7.16.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mteb
-Version: 2.7.15
+Version: 2.7.16
 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>

{mteb-2.7.15.dist-info → mteb-2.7.16.dist-info}/RECORD

@@ -1556,7 +1556,7 @@ mteb/models/model_implementations/llm2clip_models.py,sha256=X3W16uipaZ0t4Mco4lhh
 mteb/models/model_implementations/llm2vec_models.py,sha256=n86YQ8fAHU1gVtlY7tZcXq-1ab_ISxBmuk-X4MDnY4o,13348
 mteb/models/model_implementations/mcinext_models.py,sha256=T3vO9JQSmh3BICp6Y_q7j4anuA8P8LGZ4ZWnwGnF7cs,19299
 mteb/models/model_implementations/mdbr_models.py,sha256=AqsRZ-IDekIjq-FDWu0zx7Nk9ySJxaWTdRb8YhUZeu4,2828
-mteb/models/model_implementations/misc_models.py,sha256=0FkvheqPYh3JwM65F4CDlQKBDQQdjyMyfJPUdP1X2Ns,74780
+mteb/models/model_implementations/misc_models.py,sha256=JkJsyha-B5M8myLvHIwFUV14yo2lnSuBzHeO5fE9i74,73191
 mteb/models/model_implementations/mixedbread_ai_models.py,sha256=1-RD4M-16M-Rcf5CTD_R7LVoLv3cNFbmEjataQ__q94,10666
 mteb/models/model_implementations/mme5_models.py,sha256=V7BCGFkfZxkZ3ANJImvSFfP7in8OSfmkbqX-zXc_iF8,1574
 mteb/models/model_implementations/moco_models.py,sha256=6eEGpGTlI4StFRYsaNtXejhYE9GCqasUYCqB_SQy9cE,5714
@@ -1590,8 +1590,8 @@ mteb/models/model_implementations/random_baseline.py,sha256=YsITQoLbea_Iz2X84WNG
 mteb/models/model_implementations/rasgaard_models.py,sha256=_uNYP_nqJcOyoKnHNcvfJnP9gRvsv7HCWhZX2LJzQ9s,1322
 mteb/models/model_implementations/reasonir_model.py,sha256=WNWGqa9wANBL9vTdcFx51TEFXz6yHq_ygK0rij3LCL8,5217
 mteb/models/model_implementations/repllama_models.py,sha256=k6BgN2Cn41p0gQ0F1FdOTQ9OXlmFgG-2RtdvzOcCSZg,7543
-mteb/models/model_implementations/rerankers_custom.py,sha256=Bjgg_UbeHarupzzCk2rdy_Dd0_W0ZsE-DCD5v1EshnI,10953
-mteb/models/model_implementations/rerankers_monot5_based.py,sha256=6por4DPCycS8gljqKRZWUNM093bjjSVvmyQ3dzj9H6U,35321
+mteb/models/model_implementations/rerankers_custom.py,sha256=WBSA7kBRqxgb1549UwRYdtYzUovdwmW8C0PWzvGR54g,8087
+mteb/models/model_implementations/rerankers_monot5_based.py,sha256=U9ChokUEDXtkoFno-o4GeT4fXEEoFtnZn2denIafxi8,34583
 mteb/models/model_implementations/richinfoai_models.py,sha256=FsXamY-bvR5LLagtKK8fP-I5oc6B_bKp_i6_xzUYL8Y,1069
 mteb/models/model_implementations/ru_sentence_models.py,sha256=W4R985LnThJ-9XFbPnTGKb3L1QnoS3i3VXBFq94DK_w,43034
 mteb/models/model_implementations/ruri_models.py,sha256=3zYOqacB3JEnGJkMGYHqFgVkbmLo4uceJs9kzV54ivU,10819
@@ -2646,9 +2646,9 @@ mteb/types/_metadata.py,sha256=NN-W0S6a5TDV7UkpRx1pyWtGF4TyyCyoPUfHOwdeci8,2290
 mteb/types/_result.py,sha256=UKNokV9pu3G74MGebocU512aU_fFU9I9nPKnrG9Q0iE,1035
 mteb/types/_string_validators.py,sha256=PY-dYq4E8O50VS3bLYdldPWp400fl_WzUjfVSkNWe8U,523
 mteb/types/statistics.py,sha256=gElgSShKBXpfcqaZHhU_d2UHln1CyzUj8FN8KFun_UA,4087
-mteb-2.7.15.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-mteb-2.7.15.dist-info/METADATA,sha256=EoUeroRRdre5jYbplBGCJuWs-6M7cZGpzwLqSQyJKgI,14348
-mteb-2.7.15.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
-mteb-2.7.15.dist-info/entry_points.txt,sha256=8IJoEJFKoDHmVnNev-qJ9pp4Ln7_1-ma9QsXnzVCzGU,39
-mteb-2.7.15.dist-info/top_level.txt,sha256=OLVIjcQAlWBz0bdmutKlWHLF42FF0hp4uVAg3ZyiG4U,5
-mteb-2.7.15.dist-info/RECORD,,
+mteb-2.7.16.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+mteb-2.7.16.dist-info/METADATA,sha256=a-Rt1xa9ZgNdKf-JlM6EUZE_pKzEHoT6KGpFZUvnPo0,14348
+mteb-2.7.16.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
+mteb-2.7.16.dist-info/entry_points.txt,sha256=8IJoEJFKoDHmVnNev-qJ9pp4Ln7_1-ma9QsXnzVCzGU,39
+mteb-2.7.16.dist-info/top_level.txt,sha256=OLVIjcQAlWBz0bdmutKlWHLF42FF0hp4uVAg3ZyiG4U,5
+mteb-2.7.16.dist-info/RECORD,,