mteb 2.7.1__py3-none-any.whl → 2.7.3__py3-none-any.whl

mteb/models/model_implementations/no_instruct_sentence_models.py

@@ -1,15 +1,22 @@
-from collections.abc import Generator
+from __future__ import annotations
+
 from itertools import islice
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import numpy as np
 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.types import Array, BatchedInput, PromptType
+from mteb.types import PromptType
+
+if TYPE_CHECKING:
+    from collections.abc import Generator
+
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput
 
 
 # https://docs.python.org/3/library/itertools.html#itertools.batched

mteb/models/model_implementations/nomic_models.py

@@ -1,15 +1,21 @@
+from __future__ import annotations
+
 import logging
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
 import torch.nn.functional as F
 from packaging.version import Version
-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
+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
 
 logger = logging.getLogger(__name__)
 

mteb/models/model_implementations/nomic_models_vision.py

@@ -4,17 +4,18 @@ from typing import TYPE_CHECKING, Any
 
 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.types 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 import Array, BatchedInput, PromptType
 
 NOMIC_EMBED_VISION_CITATION = """@article{nussbaum2024nomicembedvision,
       title={Nomic Embed Vision: Expanding the Latent Space},

mteb/models/model_implementations/nvidia_llama_nemoretriever_colemb.py

@@ -1,14 +1,18 @@
-from typing import Any
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Any
 
 import torch
 from packaging.version import Version
 from torch.utils.data import DataLoader
 from transformers import __version__ as transformers_version
 
-from mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta
-from mteb.types import Array, BatchedInput, PromptType
+
+if TYPE_CHECKING:
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
 
 LLAMA_NEMORETRIEVER_CITATION = """@misc{xu2025llamanemoretrievercolembedtopperforming,
       title={Llama Nemoretriever Colembed: Top-Performing Text-Image Retrieval Model},

mteb/models/model_implementations/nvidia_models.py

@@ -1,11 +1,11 @@
+from __future__ import annotations
+
 import logging
-from collections.abc import Callable
-from typing import Any
+from typing import TYPE_CHECKING, Any
 
 import torch
 import torch.nn.functional as F
 from packaging.version import Version
-from torch.utils.data import DataLoader
 from tqdm import tqdm
 from transformers import AutoModel, AutoTokenizer
 from transformers import __version__ as transformers_version
@@ -16,7 +16,15 @@ from mteb.models import CrossEncoderWrapper
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.instruct_wrapper import InstructSentenceTransformerModel
 from mteb.models.model_meta import ModelMeta, ScoringFunction
-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 import TaskMetadata
+    from mteb.types import Array, BatchedInput
 
 logger = logging.getLogger(__name__)
 

mteb/models/model_implementations/octen_models.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_implementations/openai_models.py

@@ -1,15 +1,20 @@
+from __future__ import annotations
+
 import logging
-from typing import Any, ClassVar
+from typing import TYPE_CHECKING, Any, ClassVar
 
 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
+
+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/openclip_models.py

@@ -1,14 +1,19 @@
-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._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
+
+if TYPE_CHECKING:
+    from torch.utils.data import DataLoader
+
+    from mteb.abstasks.task_metadata import TaskMetadata
+    from mteb.types import Array, BatchedInput, PromptType
 
 OPENCLIP_CITATION = """@inproceedings{cherti2023reproducible,
     title={Reproducible scaling laws for contrastive language-image learning},

mteb/models/model_implementations/opensearch_neural_sparse_models.py

@@ -1,12 +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 mteb.abstasks.task_metadata import TaskMetadata
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta
-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
 
 v2_training_data = {
     "MSMARCO",

mteb/models/model_implementations/ops_moa_models.py

@@ -1,8 +1,13 @@
-import numpy as np
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
 
 from mteb.models.abs_encoder import AbsEncoder
 from mteb.models.model_meta import ModelMeta
 
+if TYPE_CHECKING:
+    from mteb.types import Array
+
 
 class OPSWrapper(AbsEncoder):
     def __init__(self, model_name: str, revision: str):
@@ -15,7 +20,7 @@ class OPSWrapper(AbsEncoder):
         )
         self.output_dim = 1536
 
-    def encode(self, sentences: list[str], **kwargs) -> np.ndarray:
+    def encode(self, sentences: list[str], **kwargs) -> Array:
         embeddings = self.model.encode(sentences, **kwargs)
         return embeddings[:, : self.output_dim]
 

mteb/models/model_implementations/pixie_models.py

@@ -0,0 +1,56 @@
+from mteb.models.model_implementations.arctic_models import (
+    ARCTIC_V2_CITATION,
+    LANGUAGES_V2_0,
+    arctic_v2_training_datasets,
+)
+from mteb.models.model_meta import (
+    ModelMeta,
+    ScoringFunction,
+)
+from mteb.models.sentence_transformer_wrapper import sentence_transformers_loader
+
+PIXIE_RUNE_V1_CITATION = """@misc{TelePIX-PIXIE-Rune-v1.0,
+  title        = {PIXIE-Rune-v1.0},
+  author       = {TelePIX AI Research Team and Bongmin Kim},
+  year         = {2026},
+  howpublished = {Hugging Face model card},
+  url          = {https://huggingface.co/telepix/PIXIE-Rune-v1.0}
+}"""
+
+PIXIE_RUNE_V1_PROMPTS = {
+    "query": "query: ",
+    "document": "",
+}
+
+# it is further fine-tuned on TelePIX proprietary IR data (not public).
+pixie_rune_v1_training_datasets = set(arctic_v2_training_datasets) | {
+    "TelePIX-Proprietary-IR-Triplets",
+}
+
+pixie_rune_v1_0 = ModelMeta(
+    loader=sentence_transformers_loader,
+    loader_kwargs={
+        "model_prompts": PIXIE_RUNE_V1_PROMPTS,
+    },
+    name="telepix/PIXIE-Rune-v1.0",
+    model_type=["dense"],
+    revision="b2486496da71191626666a88f9bfec844933a134",
+    release_date="2026-01-15",
+    languages=LANGUAGES_V2_0,
+    open_weights=True,
+    framework=["Sentence Transformers", "PyTorch", "safetensors"],
+    n_parameters=567754752,
+    memory_usage_mb=2166,
+    max_tokens=6144,
+    embed_dim=1024,
+    license="apache-2.0",
+    reference="https://huggingface.co/telepix/PIXIE-Rune-v1.0",
+    similarity_fn_name=ScoringFunction.COSINE,
+    use_instructions=True,
+    adapted_from="Snowflake/snowflake-arctic-embed-l-v2.0",
+    superseded_by=None,
+    public_training_code=None,
+    public_training_data=None,
+    training_datasets=pixie_rune_v1_training_datasets,
+    citation=PIXIE_RUNE_V1_CITATION + "\n\n" + ARCTIC_V2_CITATION,
+)

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(