mteb 2.7.2__py3-none-any.whl → 2.7.3__py3-none-any.whl

mteb/models/model_implementations/promptriever_models.py

@@ -1,15 +1,21 @@
+from __future__ import annotations
+
 import logging
-from collections.abc import Callable
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
-from torch.utils.data import DataLoader
 
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.models.models_protocols import EncoderProtocol
-from mteb.types import Array, BatchedInput, PromptType
+
+if TYPE_CHECKING:
+    from collections.abc import Callable
+
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.models.models_protocols import EncoderProtocol
+    from mteb.types import Array, BatchedInput, PromptType
 
 from .repllama_models import RepLLaMAModel, model_prompts
 

mteb/models/model_implementations/pylate_models.py

@@ -1,30 +1,36 @@
+from __future__ import annotations
+
 import heapq
 import logging
 import shutil
 import tempfile
 from pathlib import Path
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
-from torch.utils.data import DataLoader
 
 from mteb._create_dataloaders import (
     create_dataloader,
 )
 from mteb._requires_package import requires_package
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import (
-    Array,
-    BatchedInput,
-    CorpusDatasetType,
-    EncodeKwargs,
-    PromptType,
-    QueryDatasetType,
-    RetrievalOutputType,
-    TopRankedDocumentsType,
-)
+from mteb.types import PromptType
+
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import (
+        Array,
+        BatchedInput,
+        CorpusDatasetType,
+        EncodeKwargs,
+        QueryDatasetType,
+        RetrievalOutputType,
+        TopRankedDocumentsType,
+    )
+
 
 logger = logging.getLogger(__name__)
 

mteb/models/model_implementations/qwen3_models.py

@@ -1,6 +1,13 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
 from mteb.models.instruct_wrapper import InstructSentenceTransformerModel
 from mteb.models.model_meta import ModelMeta
-from mteb.models.models_protocols import EncoderProtocol, PromptType
+from mteb.types import PromptType
+
+if TYPE_CHECKING:
+    from mteb.models.models_protocols import EncoderProtocol
 
 
 def instruction_template(

mteb/models/model_implementations/random_baseline.py

@@ -5,18 +5,19 @@ from typing import TYPE_CHECKING, Any, Literal
 
 import numpy as np
 import torch
-from torch.utils.data import DataLoader
 
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.model_meta import ModelMeta
 from mteb.similarity_functions import (
     select_pairwise_similarity,
     select_similarity,
 )
-from mteb.types._encoder_io import Array, BatchedInput, PromptType
 
 if TYPE_CHECKING:
     from PIL import Image
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types._encoder_io import Array, BatchedInput, PromptType
 
 
 def _string_to_vector(text: str | None, size: int) -> np.ndarray:

mteb/models/model_implementations/repllama_models.py

@@ -1,22 +1,29 @@
+from __future__ import annotations
+
 import logging
-from collections.abc import Callable
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import numpy as np
 import torch
 import torch.nn.functional as F
-from torch.utils.data import DataLoader
 from tqdm.auto import tqdm
 
 from mteb._requires_package import requires_package
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import (
     ModelMeta,
     ScoringFunction,
 )
-from mteb.models.models_protocols import EncoderProtocol
-from mteb.types import Array, BatchedInput, PromptType
+from mteb.types import PromptType
+
+if TYPE_CHECKING:
+    from collections.abc import Callable
+
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.models.models_protocols import EncoderProtocol
+    from mteb.types import Array, BatchedInput
 
 logger = logging.getLogger(__name__)
 

mteb/models/model_implementations/rerankers_custom.py

@@ -1,16 +1,22 @@
+from __future__ import annotations
+
 import logging
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
-from torch.utils.data import DataLoader
 
 from mteb._requires_package import requires_package
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.model_meta import ModelMeta
-from mteb.types import Array, BatchedInput, PromptType
 
 from .bge_models import bge_m3_training_data
 
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
+
+
 logger = logging.getLogger(__name__)
 
 

mteb/models/model_implementations/rerankers_monot5_based.py

@@ -1,15 +1,21 @@
+from __future__ import annotations
+
 import logging
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
-from torch.utils.data import DataLoader
 
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.model_meta import ModelMeta
-from mteb.types import Array, BatchedInput, PromptType
 
 from .rerankers_custom import RerankerWrapper
 
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
+
+
 logger = logging.getLogger(__name__)
 
 

mteb/models/model_implementations/salesforce_models.py

@@ -1,12 +1,18 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
 from mteb.models.instruct_wrapper import (
     InstructSentenceTransformerModel,
     instruct_wrapper,
 )
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import PromptType
 
 from .e5_instruct import E5_MISTRAL_TRAINING_DATA
 
+if TYPE_CHECKING:
+    from mteb.types import PromptType
+
 
 def instruction_template(
     instruction: str, prompt_type: PromptType | None = None

mteb/models/model_implementations/seed_1_6_embedding_models.py

@@ -13,16 +13,18 @@ import torch
 from torch.utils.data import DataLoader
 
 from mteb._requires_package import requires_package
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_implementations.bge_models import bge_chinese_training_data
 from mteb.models.model_implementations.nvidia_models import nvidia_training_datasets
 from mteb.models.model_meta import ModelMeta
-from mteb.types import Array, BatchedInput, PromptType
+from mteb.types import PromptType
 
 if TYPE_CHECKING:
     from PIL import Image
 
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput
+
 
 logger = logging.getLogger(__name__)
 

mteb/models/model_implementations/seed_1_6_embedding_models_1215.py

@@ -15,15 +15,18 @@ from torch.utils.data import DataLoader
 from tqdm import tqdm
 
 from mteb._requires_package import requires_package
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_implementations.bge_models import bge_chinese_training_data
 from mteb.models.model_implementations.nvidia_models import nvidia_training_datasets
 from mteb.models.model_meta import ModelMeta
-from mteb.types import Array, BatchedInput, PromptType
+from mteb.types import PromptType
 
 if TYPE_CHECKING:
     from PIL import Image
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput
 
 
 logger = logging.getLogger(__name__)

mteb/models/model_implementations/seed_models.py

@@ -9,7 +9,7 @@ from tqdm.auto import tqdm
 from mteb._requires_package import requires_package
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta
-from mteb.models.models_protocols import PromptType
+from mteb.types import PromptType
 
 from .bge_models import bge_chinese_training_data
 from .nvidia_models import nvidia_training_datasets

mteb/models/model_implementations/siglip_models.py

@@ -1,13 +1,18 @@
-from typing import Any
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Any
 
 import torch
-from torch.utils.data import DataLoader
 from tqdm.auto import tqdm
 
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import Array, BatchedInput, PromptType
+
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
 
 SIGLIP_CITATION = """@misc{zhai2023sigmoid,
       title={Sigmoid Loss for Language Image Pre-Training},

mteb/models/model_implementations/slm_models.py

@@ -13,24 +13,27 @@ Based on:
 from __future__ import annotations
 
 import logging
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
-from torch.utils.data import DataLoader
 from tqdm.auto import tqdm
 
 from mteb._requires_package import (
     requires_image_dependencies,
     requires_package,
 )
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_implementations.colpali_models import (
     COLPALI_CITATION,
     COLPALI_TRAINING_DATA,
 )
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import Array, BatchedInput, PromptType
+
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
 
 logger = logging.getLogger(__name__)
 

mteb/models/model_implementations/uae_models.py

@@ -1,13 +1,18 @@
+from __future__ import annotations
+
 import logging
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
-from torch.utils.data import DataLoader
 
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.model_meta import ModelMeta, ScoringFunction
 from mteb.models.sentence_transformer_wrapper import SentenceTransformerEncoderWrapper
-from mteb.types import Array, BatchedInput, PromptType
+
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
 
 logger = logging.getLogger(__name__)
 

mteb/models/model_implementations/vdr_models.py

@@ -1,6 +1,12 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
 from mteb.models.instruct_wrapper import InstructSentenceTransformerModel
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import PromptType
+
+if TYPE_CHECKING:
+    from mteb.types import PromptType
 
 
 def instruction_template(

mteb/models/model_implementations/vista_models.py

@@ -1,14 +1,19 @@
-from typing import Any, Literal
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Any, Literal
 
 import torch
-from torch.utils.data import DataLoader
 from tqdm.auto import tqdm
 
 from mteb._requires_package import requires_image_dependencies
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import Array, BatchedInput, PromptType
+
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
 
 VISTA_CITATION = """@article{zhou2024vista,
   title={VISTA: Visualized Text Embedding For Universal Multi-Modal Retrieval},

mteb/models/model_implementations/vlm2vec_models.py

@@ -1,8 +1,9 @@
+from __future__ import annotations
+
 import logging
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
-from torch.utils.data import DataLoader
 from tqdm.auto import tqdm
 
 from mteb._requires_package import (
@@ -10,10 +11,14 @@ from mteb._requires_package import (
     requires_package,
     suggest_package,
 )
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import Array, BatchedInput, PromptType
+
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
 
 logger = logging.getLogger(__name__)
 

mteb/models/model_implementations/voyage_models.py

@@ -1,16 +1,22 @@
+from __future__ import annotations
+
 import time
 from functools import wraps
-from typing import Any, Literal
+from typing import TYPE_CHECKING, Any, Literal
 
 import numpy as np
-from torch.utils.data import DataLoader
 from tqdm.auto import tqdm
 
 from mteb._requires_package import requires_package
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import Array, BatchedInput, PromptType
+from mteb.types import PromptType
+
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput
 
 VOYAGE_TRAINING_DATA = set(
     # Self-reported (message from VoyageAI member)

mteb/models/model_implementations/voyage_v.py

@@ -4,17 +4,19 @@ import logging
 from typing import TYPE_CHECKING, Any, Literal
 
 import torch
-from torch.utils.data import DataLoader
 from tqdm.auto import tqdm
 
 from mteb._requires_package import requires_image_dependencies, requires_package
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-from mteb.types import Array, BatchedInput, PromptType
+from mteb.types import PromptType
 
 if TYPE_CHECKING:
     from PIL import Image
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput
 
 logger = logging.getLogger(__name__)
 
@@ -27,6 +29,8 @@ def _downsample_image(
     Returns:
         The downsampled image.
     """
+    from PIL.Image import Resampling
+
     width, height = image.size
     pixels = width * height
 
@@ -42,15 +46,15 @@ def _downsample_image(
         logger.info(
             f"Downsampling image from {width}x{height} to {new_width}x{new_height}"
         )
-        return image.resize(new_size, Image.LANCZOS)
+        return image.resize(new_size, Resampling.LANCZOS)
     if width > height:
         if width > 10000:
             logger.error("Processing extremely wide images.")
-            return image.resize((10000, height), Image.LANCZOS)
+            return image.resize((10000, height), Resampling.LANCZOS)
     else:
         if height > 10000:
             logger.error("Processing extremely high images.")
-            return image.resize((width, 10000), Image.LANCZOS)
+            return image.resize((width, 10000), Resampling.LANCZOS)
     return image
 
 

mteb/models/model_implementations/yuan_models_en.py

@@ -1,6 +1,6 @@
 from mteb.models.instruct_wrapper import InstructSentenceTransformerModel
 from mteb.models.model_meta import ModelMeta
-from mteb.models.models_protocols import PromptType
+from mteb.types import PromptType
 
 
 def instruction_template(

mteb/models/model_meta.py

@@ -3,7 +3,7 @@ from __future__ import annotations
 import json
 import logging
 import warnings
-from collections.abc import Callable, Sequence
+from collections.abc import Callable
 from dataclasses import field
 from enum import Enum
 from functools import partial
@@ -11,9 +11,7 @@ from pathlib import Path
 from typing import TYPE_CHECKING, Any, Literal, cast
 
 from huggingface_hub import (
-    GitCommitInfo,
     ModelCard,
-    ModelCardData,
     get_safetensors_metadata,
     hf_hub_download,
     list_repo_commits,
@@ -30,17 +28,24 @@ from huggingface_hub.errors import (
 )
 from pydantic import BaseModel, ConfigDict, field_validator, model_validator
 from transformers import AutoConfig
-from typing_extensions import Self
 
 from mteb._helpful_enum import HelpfulStrEnum
 from mteb.languages import check_language_code
-from mteb.models.models_protocols import EncoderProtocol, MTEBModels
+from mteb.models.models_protocols import MTEBModels
 from mteb.types import ISOLanguageScript, Licenses, Modalities, StrDate, StrURL
 
 if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+    from huggingface_hub import (
+        GitCommitInfo,
+        ModelCardData,
+    )
     from sentence_transformers import CrossEncoder, SentenceTransformer
+    from typing_extensions import Self
 
     from mteb.abstasks import AbsTask
+    from mteb.models.models_protocols import EncoderProtocol
 
 
 logger = logging.getLogger(__name__)
@@ -479,7 +484,7 @@ class ModelMeta(BaseModel):
         if isinstance(tasks[0], str):
             benchmark_datasets = set(tasks)
         else:
-            tasks = cast(Sequence["AbsTask"], tasks)
+            tasks = cast("Sequence[AbsTask]", tasks)
             benchmark_datasets = set()
             for task in tasks:
                 benchmark_datasets.add(task.metadata.name)
@@ -534,7 +539,7 @@ class ModelMeta(BaseModel):
         if isinstance(tasks[0], str):
             benchmark_datasets = set(tasks)
         else:
-            tasks = cast(Sequence["AbsTask"], tasks)
+            tasks = cast("Sequence[AbsTask]", tasks)
             benchmark_datasets = {task.metadata.name for task in tasks}
         overlap = training_datasets & benchmark_datasets
         perc_overlap = 100 * (len(overlap) / len(benchmark_datasets))

mteb/models/models_protocols.py

@@ -1,22 +1,23 @@
-from typing import TYPE_CHECKING, Any, Protocol, runtime_checkable
+from __future__ import annotations
 
-from torch.utils.data import DataLoader
-from typing_extensions import Unpack
-
-from mteb.abstasks.task_metadata import TaskMetadata
-from mteb.types import (
-    Array,
-    BatchedInput,
-    CorpusDatasetType,
-    EncodeKwargs,
-    PromptType,
-    QueryDatasetType,
-    RetrievalOutputType,
-    TopRankedDocumentsType,
-)
+from typing import TYPE_CHECKING, Any, Protocol, runtime_checkable
 
 if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+    from typing_extensions import Unpack
+
+    from mteb.abstasks.task_metadata import TaskMetadata
     from mteb.models.model_meta import ModelMeta
+    from mteb.types import (
+        Array,
+        BatchedInput,
+        CorpusDatasetType,
+        EncodeKwargs,
+        PromptType,
+        QueryDatasetType,
+        RetrievalOutputType,
+        TopRankedDocumentsType,
+    )
 
 
 @runtime_checkable
@@ -72,7 +73,7 @@ class SearchProtocol(Protocol):
         ...
 
     @property
-    def mteb_model_meta(self) -> "ModelMeta":
+    def mteb_model_meta(self) -> ModelMeta:
         """Metadata of the model"""
         ...
 
@@ -177,7 +178,7 @@ class EncoderProtocol(Protocol):
         ...
 
     @property
-    def mteb_model_meta(self) -> "ModelMeta":
+    def mteb_model_meta(self) -> ModelMeta:
         """Metadata of the model"""
         ...
 
@@ -236,7 +237,7 @@ class CrossEncoderProtocol(Protocol):
         ...
 
     @property
-    def mteb_model_meta(self) -> "ModelMeta":
+    def mteb_model_meta(self) -> ModelMeta:
         """Metadata of the model"""
         ...
 

mteb/models/search_encoder_index/search_backend_protocol.py

@@ -1,7 +1,11 @@
-from collections.abc import Callable
-from typing import Protocol
+from __future__ import annotations
 
-from mteb.types import Array, TopRankedDocumentsType
+from typing import TYPE_CHECKING, Protocol
+
+if TYPE_CHECKING:
+    from collections.abc import Callable
+
+    from mteb.types import Array, TopRankedDocumentsType
 
 
 class IndexEncoderSearchProtocol(Protocol):