mteb 2.7.2__py3-none-any.whl → 2.7.4__py3-none-any.whl

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__)
 
@@ -346,6 +352,7 @@ colbert_v2 = ModelMeta(
     public_training_data=None,
     release_date="2024-09-21",
     n_parameters=int(110 * 1e6),
+    n_embedding_parameters=23_440_896,
     memory_usage_mb=418,
     max_tokens=180,
     embed_dim=None,
@@ -402,6 +409,7 @@ jina_colbert_v2 = ModelMeta(
     public_training_data=None,
     release_date="2024-08-16",
     n_parameters=int(559 * 1e6),
+    n_embedding_parameters=None,
     memory_usage_mb=1067,
     max_tokens=8192,
     embed_dim=None,
@@ -458,6 +466,7 @@ lightonai__gte_moderncolbert_v1 = ModelMeta(
     public_training_data="https://huggingface.co/datasets/lightonai/ms-marco-en-bge-gemma",
     release_date="2025-04-30",
     n_parameters=int(149 * 1e6),
+    n_embedding_parameters=None,
     memory_usage_mb=None,
     max_tokens=8192,
     embed_dim=None,

mteb/models/model_implementations/qodo_models.py

@@ -36,6 +36,7 @@ Qodo_Embed_1_1_5B = ModelMeta(
     revision="84bbef079b32e8823ec226d4e9e92902706b0eb6",
     release_date="2025-02-19",
     n_parameters=1_780_000_000,
+    n_embedding_parameters=232_928_256,
     memory_usage_mb=6776,
     embed_dim=1536,
     license="https://huggingface.co/Qodo/Qodo-Embed-1-1.5B/blob/main/LICENSE",
@@ -59,6 +60,7 @@ Qodo_Embed_1_7B = ModelMeta(
     revision="f9edd9bf7f687c0e832424058e265120f603cd81",
     release_date="2025-02-24",
     n_parameters=7_613_000_000,
+    n_embedding_parameters=None,
     memory_usage_mb=29040,
     embed_dim=3584,
     license="https://huggingface.co/Qodo/Qodo-Embed-1-1.5B/blob/main/LICENSE",

mteb/models/model_implementations/qtack_models.py

@@ -31,6 +31,7 @@ mini_gte = ModelMeta(
     revision="7fbe6f9b4cc42615e0747299f837ad7769025492",
     release_date="2025-01-28",
     n_parameters=int(66.3 * 1e6),
+    n_embedding_parameters=23_440_896,
     memory_usage_mb=253,
     embed_dim=768,
     license="apache-2.0",

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(
@@ -140,6 +147,7 @@ Qwen3_Embedding_0B6 = ModelMeta(
     revision="b22da495047858cce924d27d76261e96be6febc0",  # Commit of @tomaarsen
     release_date="2025-06-05",
     n_parameters=595776512,
+    n_embedding_parameters=None,
     memory_usage_mb=1136,
     embed_dim=1024,
     max_tokens=32768,
@@ -163,6 +171,7 @@ Qwen3_Embedding_4B = ModelMeta(
     revision="636cd9bf47d976946cdbb2b0c3ca0cb2f8eea5ff",  # Commit of @tomaarsen
     release_date="2025-06-05",
     n_parameters=4021774336,
+    n_embedding_parameters=None,
     memory_usage_mb=7671,
     embed_dim=2560,
     max_tokens=32768,
@@ -186,6 +195,7 @@ Qwen3_Embedding_8B = ModelMeta(
     revision="4e423935c619ae4df87b646a3ce949610c66241c",  # Commit of @tomaarsen
     release_date="2025-06-05",
     n_parameters=7567295488,
+    n_embedding_parameters=None,
     memory_usage_mb=14433,
     embed_dim=4096,
     max_tokens=32768,

mteb/models/model_implementations/qzhou_models.py

@@ -64,6 +64,7 @@ QZhou_Embedding = ModelMeta(
     revision="f1e6c03ee3882e7b9fa5cec91217715272e433b8",
     release_date="2025-08-24",
     n_parameters=7_070_619_136,
+    n_embedding_parameters=None,
     memory_usage_mb=14436,
     embed_dim=3584,
     license="apache-2.0",
@@ -98,6 +99,7 @@ QZhou_Embedding_Zh = ModelMeta(
     revision="0321ccb126413d1e49c5ce908e802b63d35f18e2",
     release_date="2025-09-28",
     n_parameters=7_575_747_328,
+    n_embedding_parameters=None,
     memory_usage_mb=29431,
     embed_dim=1792,
     license="apache-2.0",

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/rasgaard_models.py

@@ -12,6 +12,7 @@ potion_base_8m = ModelMeta(
     revision="387897cfb09992e6d45ea9cd7b28b9fcf119e23a",
     release_date="2025-10-08",
     n_parameters=22893312,
+    n_embedding_parameters=22893312,
     memory_usage_mb=87,
     max_tokens=np.inf,
     embed_dim=256,

mteb/models/model_implementations/reasonir_model.py

@@ -36,12 +36,76 @@ REASONIR_TRAINING_DATA = {
     "DuRetrieval",
     "QuoraRetrieval",
 }
+_prompts_dict = {
+    "BrightBiologyRetrieval": {
+        "query": "Given a Biology post, retrieve relevant passages that help answer the post"
+    },
+    "BrightEarthScienceRetrieval": {
+        "query": "Given a Earth Science post, retrieve relevant passages that help answer the post"
+    },
+    "BrightEconomicsRetrieval": {
+        "query": "Given a Economics post, retrieve relevant passages that help answer the post"
+    },
+    "BrightPsychologyRetrieval": {
+        "query": "Given a Psychology post, retrieve relevant passages that help answer the post"
+    },
+    "BrightRoboticsRetrieval": {
+        "query": "Given a Robotics post, retrieve relevant passages that help answer the post"
+    },
+    "BrightStackoverflowRetrieval": {
+        "query": "Given a Stackoverflow post, retrieve relevant passages that help answer the post"
+    },
+    "BrightSustainableLivingRetrieval": {
+        "query": "Given a Sustainable Living post, retrieve relevant passages that help answer the post"
+    },
+    "BrightPonyRetrieval": {
+        "query": "Given a Pony question, retrieve relevant passages that help answer the question"
+    },
+    "BrightLeetcodeRetrieval": {
+        "query": "Given a coding problem, retrieve relevant examples that help answer the problem",
+    },
+    "BrightAopsRetrieval": {
+        "query": "Given a Math problem, retrieve relevant examples that help answer the problem"
+    },
+    "BrightTheoremQATheoremsRetrieval": {
+        "query": "Given a Math problem, retrieve relevant theorems that help answer the problem",
+    },
+    "BrightTheoremQAQuestionsRetrieval": {
+        "query": "Given a Math problem, retrieve relevant examples that help answer the problem",
+    },
+    "BrightBiologyLongRetrieval": {
+        "query": "Given a Biology post, retrieve relevant documents that help answer the post"
+    },
+    "BrightEarthScienceLongRetrieval": {
+        "query": "Given a Earth Science post, retrieve relevant documents that help answer the post"
+    },
+    "BrightEconomicsLongRetrieval": {
+        "query": "Given a Economics post, retrieve relevant documents that help answer the post"
+    },
+    "BrightPsychologyLongRetrieval": {
+        "query": "Given a Psychology post, retrieve relevant documents that help answer the post"
+    },
+    "BrightRoboticsLongRetrieval": {
+        "query": "Given a Robotics post, retrieve relevant documents that help answer the post"
+    },
+    "BrightStackoverflowLongRetrieval": {
+        "query": "Given a Stackoverflow post, retrieve relevant documents that help answer the post"
+    },
+    "BrightSustainableLivingLongRetrieval": {
+        "query": "Given a Sustainable Living post, retrieve relevant documents that help answer the post"
+    },
+    "BrightPonyLongRetrieval": {
+        "query": "Given a Pony question, retrieve relevant documents that help answer the question"
+    },
+}
+
 
 ReasonIR_8B = ModelMeta(
     loader=InstructSentenceTransformerModel,
     loader_kwargs=dict(
         instruction_template=instruction_template,
         trust_remote_code=True,
+        prompts_dict=_prompts_dict,
     ),
     name="ReasonIR/ReasonIR-8B",
     model_type=["dense"],
@@ -50,6 +114,7 @@ ReasonIR_8B = ModelMeta(
     revision="c3d0690370ff4a8c3d3882d8dfa85c43650034fa",
     release_date="2025-04-29",
     n_parameters=7_500_000_000,
+    n_embedding_parameters=None,
     memory_usage_mb=None,
     embed_dim=4096,
     license="cc-by-nc-4.0",

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__)
 
@@ -172,6 +179,7 @@ repllama_llama2_original = ModelMeta(
         "mMARCO-NL",  # translation not trained on
     },
     n_parameters=7_000_000,
+    n_embedding_parameters=131_072_000,
     memory_usage_mb=27,
     max_tokens=4096,
     embed_dim=4096,
@@ -201,6 +209,7 @@ repllama_llama2_reproduced = ModelMeta(
     revision="01c7f73d771dfac7d292323805ebc428287df4f9-ad5c1d0938a1e02954bcafb4d811ba2f34052e71",  # base-peft revision
     release_date="2024-09-15",
     n_parameters=7_000_000,
+    n_embedding_parameters=None,
     memory_usage_mb=27,
     max_tokens=4096,
     embed_dim=4096,

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__)
 
 
@@ -225,6 +231,7 @@ monobert_large = ModelMeta(
     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,
@@ -250,6 +257,7 @@ jina_reranker_multilingual = ModelMeta(
     revision="126747772a932960028d9f4dc93bd5d9c4869be4",
     release_date="2024-09-26",
     n_parameters=None,
+    n_embedding_parameters=None,
     memory_usage_mb=531,
     max_tokens=None,
     embed_dim=None,
@@ -313,6 +321,7 @@ bge_reranker_v2_m3 = ModelMeta(
     revision="953dc6f6f85a1b2dbfca4c34a2796e7dde08d41e",
     release_date="2024-06-24",
     n_parameters=None,
+    n_embedding_parameters=256_002_048,
     memory_usage_mb=2166,
     max_tokens=None,
     embed_dim=None,

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__)
 
 
@@ -321,6 +327,7 @@ monot5_small = ModelMeta(
     revision="77f8e3f7b1eb1afe353aa21a7c3a2fc8feca702e",
     release_date="2022-03-28",
     n_parameters=None,
+    n_embedding_parameters=16_449_536,
     memory_usage_mb=None,
     max_tokens=None,
     embed_dim=None,
@@ -363,6 +370,7 @@ monot5_base = ModelMeta(
       url={https://arxiv.org/abs/2206.02873},
     }""",
     n_parameters=None,
+    n_embedding_parameters=24_674_304,
     memory_usage_mb=None,
     max_tokens=None,
     embed_dim=None,
@@ -387,6 +395,7 @@ monot5_large = ModelMeta(
     revision="48cfad1d8dd587670393f27ee8ec41fde63e3d98",
     release_date="2022-03-28",
     n_parameters=None,
+    n_embedding_parameters=32_899_072,
     memory_usage_mb=None,
     max_tokens=None,
     embed_dim=None,
@@ -420,6 +429,7 @@ monot5_3b = ModelMeta(
     revision="bc0c419a438c81f592f878ce32430a1823f5db6c",
     release_date="2022-03-28",
     n_parameters=None,
+    n_embedding_parameters=32_899_072,
     memory_usage_mb=None,
     max_tokens=None,
     embed_dim=None,
@@ -476,6 +486,7 @@ flant5_base = ModelMeta(
         # "qed": ["train"],
     ),
     n_parameters=None,
+    n_embedding_parameters=24_674_304,
     memory_usage_mb=944,
     max_tokens=None,
     embed_dim=None,
@@ -522,6 +533,7 @@ flant5_large = ModelMeta(
         # "qed": ["train"],
     ),
     n_parameters=None,
+    n_embedding_parameters=32_899_072,
     memory_usage_mb=2987,
     max_tokens=None,
     embed_dim=None,
@@ -568,6 +580,7 @@ flant5_xl = ModelMeta(
         # "qed": ["train"],
     ),
     n_parameters=None,
+    n_embedding_parameters=65_798_144,
     memory_usage_mb=10871,
     max_tokens=None,
     embed_dim=None,
@@ -614,6 +627,7 @@ flant5_xxl = ModelMeta(
         # "qed": ["train"],
     ),
     n_parameters=None,
+    n_embedding_parameters=131_596_288,
     memory_usage_mb=42980,
     max_tokens=None,
     embed_dim=None,
@@ -638,6 +652,7 @@ llama2_7b = ModelMeta(
     revision="01c7f73d771dfac7d292323805ebc428287df4f9",
     release_date="2023-07-18",
     n_parameters=None,
+    n_embedding_parameters=131_072_000,
     memory_usage_mb=None,
     max_tokens=None,
     embed_dim=None,
@@ -680,6 +695,7 @@ llama2_7b_chat = ModelMeta(
       url={https://arxiv.org/abs/2307.09288},
     }""",
     n_parameters=None,
+    n_embedding_parameters=131_072_000,
     memory_usage_mb=None,
     max_tokens=None,
     embed_dim=None,
@@ -704,6 +720,7 @@ mistral_7b = ModelMeta(
     revision="3ad372fc79158a2148299e3318516c786aeded6c",
     release_date="2023-12-11",
     n_parameters=None,
+    n_embedding_parameters=None,
     memory_usage_mb=None,
     max_tokens=None,
     embed_dim=None,
@@ -740,6 +757,7 @@ followir_7b = ModelMeta(
         # "jhu-clsp/FollowIR-train"
     ),
     n_parameters=None,
+    n_embedding_parameters=None,
     memory_usage_mb=13813,
     max_tokens=None,
     embed_dim=None,
@@ -890,6 +908,7 @@ mt5_base_mmarco_v2 = ModelMeta(
     """,
     training_datasets={"MSMARCO"},
     n_parameters=None,
+    n_embedding_parameters=192_086_016,
     memory_usage_mb=None,
     max_tokens=None,
     embed_dim=None,
@@ -913,6 +932,7 @@ mt5_13b_mmarco_100k = ModelMeta(
     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,

mteb/models/model_implementations/richinfoai_models.py

@@ -15,6 +15,7 @@ ritrieve_zh_v1 = ModelMeta(
     revision="f8d5a707656c55705027678e311f9202c8ced12c",
     release_date="2025-03-25",
     n_parameters=int(326 * 1e6),
+    n_embedding_parameters=21_635_072,
     memory_usage_mb=1242,
     embed_dim=1792,
     license="mit",

mteb/models/model_implementations/ru_sentence_models.py

@@ -244,6 +244,7 @@ rubert_tiny = ModelMeta(
     revision="5441c5ea8026d4f6d7505ec004845409f1259fb1",
     release_date="2021-05-24",
     n_parameters=11_900_000,
+    n_embedding_parameters=9_223_968,
     memory_usage_mb=45,
     embed_dim=312,
     license="mit",
@@ -270,6 +271,7 @@ rubert_tiny2 = ModelMeta(
     revision="dad72b8f77c5eef6995dd3e4691b758ba56b90c3",
     release_date="2021-10-28",
     n_parameters=29_400_000,
+    n_embedding_parameters=26_154_336,
     memory_usage_mb=112,
     embed_dim=312,
     license="mit",
@@ -297,6 +299,7 @@ sbert_large_nlu_ru = ModelMeta(
     revision="af977d5dfa46a3635e29bf0ef383f2df2a08d47a",
     release_date="2020-11-20",
     n_parameters=427_000_000,
+    n_embedding_parameters=123_021_312,
     memory_usage_mb=1629,
     embed_dim=1024,
     license="mit",
@@ -323,6 +326,7 @@ sbert_large_mt_nlu_ru = ModelMeta(
     revision="05300876c2b83f46d3ddd422a7f17e45cf633bb0",
     release_date="2021-05-18",
     n_parameters=427_000_000,
+    n_embedding_parameters=123_021_312,
     memory_usage_mb=1629,
     embed_dim=1024,
     license="not specified",
@@ -351,6 +355,7 @@ user_base_ru = ModelMeta(
     revision="436a489a2087d61aa670b3496a9915f84e46c861",
     release_date="2024-06-10",
     n_parameters=427_000_000,
+    n_embedding_parameters=38_603_520,
     memory_usage_mb=473,
     embed_dim=768,
     license="apache-2.0",
@@ -412,6 +417,7 @@ user_bge_m3 = ModelMeta(
     revision="0cc6cfe48e260fb0474c753087a69369e88709ae",
     release_date="2024-07-05",
     n_parameters=359_026_688,
+    n_embedding_parameters=47_273_984,
     memory_usage_mb=1370,
     embed_dim=1024,
     license="apache-2.0",
@@ -463,6 +469,7 @@ deberta_v1_ru = ModelMeta(
     revision="bdd30b0e19757e6940c92c7aff19e8fc0a60dff4",
     release_date="2023-02-07",
     n_parameters=124_000_000,
+    n_embedding_parameters=38_603_520,
     memory_usage_mb=473,
     embed_dim=768,
     license="apache-2.0",
@@ -494,6 +501,7 @@ rubert_base_cased = ModelMeta(
     revision="4036cab694767a299f2b9e6492909664d9414229",
     release_date="2020-03-04",
     n_parameters=1280_000_000,
+    n_embedding_parameters=91_812_096,
     memory_usage_mb=4883,
     embed_dim=768,
     license="not specified",
@@ -530,6 +538,7 @@ distilrubert_small_cased_conversational = ModelMeta(
     revision="e348066b4a7279b97138038299bddc6580a9169a",
     release_date="2022-06-28",
     n_parameters=107_000_000,
+    n_embedding_parameters=91_812_096,
     memory_usage_mb=408,
     embed_dim=768,
     license="not specified",
@@ -565,6 +574,7 @@ rubert_base_cased_sentence = ModelMeta(
     revision="78b5122d6365337dd4114281b0d08cd1edbb3bc8",
     release_date="2020-03-04",
     n_parameters=107_000_000,
+    n_embedding_parameters=91_812_096,
     memory_usage_mb=408,
     embed_dim=768,
     license="not specified",
@@ -590,6 +600,7 @@ labse_en_ru = ModelMeta(
     revision="cf0714e606d4af551e14ad69a7929cd6b0da7f7e",
     release_date="2021-06-10",
     n_parameters=129_000_000,
+    n_embedding_parameters=42_303_744,
     memory_usage_mb=492,
     embed_dim=768,
     license="not specified",
@@ -618,6 +629,7 @@ rubert_tiny_turbo = ModelMeta(
     revision="8ce0cf757446ce9bb2d5f5a4ac8103c7a1049054",
     release_date="2024-06-21",
     n_parameters=29_200_000,
+    n_embedding_parameters=26_154_336,
     memory_usage_mb=111,
     embed_dim=312,
     license="mit",
@@ -641,6 +653,7 @@ rubert_mini_frida = ModelMeta(
     revision="19b279b78afd945b5ccae78f63e284909814adc2",
     release_date="2025-03-02",
     n_parameters=32_300_000,
+    n_embedding_parameters=26_154_336,
     memory_usage_mb=123,
     embed_dim=312,
     license="mit",
@@ -669,6 +682,7 @@ labse_ru_turbo = ModelMeta(
     revision="1940b046c6b5e125df11722b899130329d0a46da",
     release_date="2024-06-27",
     n_parameters=129_000_000,
+    n_embedding_parameters=42_303_744,
     memory_usage_mb=490,
     embed_dim=768,
     license="mit",
@@ -720,6 +734,7 @@ rosberta_ru_en = ModelMeta(
     use_instructions=True,
     reference="https://huggingface.co/ai-forever/ru-en-RoSBERTa",
     n_parameters=404_000_000,
+    n_embedding_parameters=100_869_120,
     memory_usage_mb=1540,
     max_tokens=512,
     embed_dim=1024,
@@ -886,6 +901,7 @@ frida = ModelMeta(
     use_instructions=True,
     reference="https://huggingface.co/ai-forever/FRIDA",
     n_parameters=823_000_000,
+    n_embedding_parameters=143_847_936,
     memory_usage_mb=3141,
     max_tokens=512,
     embed_dim=1536,
@@ -918,6 +934,7 @@ giga_embeddings = ModelMeta(
     revision="0ad5b29bfecd806cecc9d66b927d828a736594dc",
     release_date="2025-09-23",
     n_parameters=3_227_176_961,
+    n_embedding_parameters=None,
     memory_usage_mb=12865,
     embed_dim=2048,
     license="mit",
@@ -950,6 +967,7 @@ berta = ModelMeta(
     revision="914c8c8aed14042ed890fc2c662d5e9e66b2faa7",
     release_date="2025-03-10",
     n_parameters=128_000_000,
+    n_embedding_parameters=42_303_744,
     memory_usage_mb=489,
     embed_dim=768,
     license="mit",
@@ -1025,6 +1043,7 @@ user2_small = ModelMeta(
     use_instructions=True,
     reference="https://huggingface.co/collections/deepvk/user2-6802650d7210f222ec60e05f",
     n_parameters=34_400_000,
+    n_embedding_parameters=None,
     memory_usage_mb=131,
     max_tokens=8192,
     embed_dim=384,
@@ -1058,6 +1077,7 @@ user2_base = ModelMeta(
     use_instructions=True,
     reference="https://huggingface.co/collections/deepvk/user2-6802650d7210f222ec60e05f",
     n_parameters=149_000_000,
+    n_embedding_parameters=None,
     memory_usage_mb=568,
     max_tokens=8192,
     embed_dim=768,