mteb 2.5.1__py3-none-any.whl → 2.5.3__py3-none-any.whl

mteb/models/model_implementations/rerankers_custom.py

@@ -219,6 +219,7 @@ monobert_large = ModelMeta(
         fp_options="float16",
     ),
     name="castorini/monobert-large-msmarco",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="0a97706f3827389da43b83348d5d18c9d53876fa",
@@ -234,7 +235,6 @@ monobert_large = ModelMeta(
     use_instructions=None,
     training_datasets=None,
     framework=["Sentence Transformers", "PyTorch"],
-    is_cross_encoder=True,
 )
 
 # languages unclear: https://huggingface.co/jinaai/jina-reranker-v2-base-multilingual/discussions/28
@@ -244,6 +244,7 @@ jina_reranker_multilingual = ModelMeta(
         fp_options="float16",
     ),
     name="jinaai/jina-reranker-v2-base-multilingual",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="126747772a932960028d9f4dc93bd5d9c4869be4",
@@ -259,7 +260,6 @@ jina_reranker_multilingual = ModelMeta(
     use_instructions=None,
     training_datasets=None,
     framework=["Sentence Transformers", "PyTorch"],
-    is_cross_encoder=True,
 )
 
 bge_reranker_v2_m3 = ModelMeta(
@@ -268,6 +268,7 @@ bge_reranker_v2_m3 = ModelMeta(
         fp_options="float16",
     ),
     name="BAAI/bge-reranker-v2-m3",
+    model_type=["cross-encoder"],
     languages=[
         "eng-Latn",
         "ara-Arab",
@@ -316,7 +317,6 @@ bge_reranker_v2_m3 = ModelMeta(
     use_instructions=None,
     training_datasets=bge_m3_training_data,
     framework=["Sentence Transformers", "PyTorch"],
-    is_cross_encoder=True,
     citation="""
     @misc{li2023making,
       title={Making Large Language Models A Better Foundation For Dense Retrieval},

mteb/models/model_implementations/rerankers_monot5_based.py

@@ -315,6 +315,7 @@ monot5_small = ModelMeta(
         fp_options="float16",
     ),
     name="castorini/monot5-small-msmarco-10k",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="77f8e3f7b1eb1afe353aa21a7c3a2fc8feca702e",
@@ -330,7 +331,6 @@ monot5_small = ModelMeta(
     use_instructions=None,
     training_datasets=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
     citation="""@misc{rosa2022parameterleftbehinddistillation,
       title={No Parameter Left Behind: How Distillation and Model Size Affect Zero-Shot Retrieval},
       author={Guilherme Moraes Rosa and Luiz Bonifacio and Vitor Jeronymo and Hugo Abonizio and Marzieh Fadaee and Roberto Lotufo and Rodrigo Nogueira},
@@ -348,6 +348,7 @@ monot5_base = ModelMeta(
         fp_options="float16",
     ),
     name="castorini/monot5-base-msmarco-10k",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="f15657ab3d2a5dd0b9a30c8c0b6a0a73c9cb5884",
@@ -372,7 +373,6 @@ monot5_base = ModelMeta(
     use_instructions=None,
     training_datasets=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
 )
 
 monot5_large = ModelMeta(
@@ -381,6 +381,7 @@ monot5_large = ModelMeta(
         fp_options="float16",
     ),
     name="castorini/monot5-large-msmarco-10k",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="48cfad1d8dd587670393f27ee8ec41fde63e3d98",
@@ -396,7 +397,6 @@ monot5_large = ModelMeta(
     use_instructions=None,
     training_datasets=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
     citation="""@misc{rosa2022parameterleftbehinddistillation,
       title={No Parameter Left Behind: How Distillation and Model Size Affect Zero-Shot Retrieval},
       author={Guilherme Moraes Rosa and Luiz Bonifacio and Vitor Jeronymo and Hugo Abonizio and Marzieh Fadaee and Roberto Lotufo and Rodrigo Nogueira},
@@ -414,6 +414,7 @@ monot5_3b = ModelMeta(
         fp_options="float16",
     ),
     name="castorini/monot5-3b-msmarco-10k",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="bc0c419a438c81f592f878ce32430a1823f5db6c",
@@ -429,7 +430,6 @@ monot5_3b = ModelMeta(
     use_instructions=None,
     training_datasets=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
     citation="""@misc{rosa2022parameterleftbehinddistillation,
       title={No Parameter Left Behind: How Distillation and Model Size Affect Zero-Shot Retrieval},
       author={Guilherme Moraes Rosa and Luiz Bonifacio and Vitor Jeronymo and Hugo Abonizio and Marzieh Fadaee and Roberto Lotufo and Rodrigo Nogueira},
@@ -447,6 +447,7 @@ flant5_base = ModelMeta(
         fp_options="float16",
     ),
     name="google/flan-t5-base",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="7bcac572ce56db69c1ea7c8af255c5d7c9672fc2",
@@ -484,7 +485,6 @@ flant5_base = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
 )
 
 flant5_large = ModelMeta(
@@ -493,6 +493,7 @@ flant5_large = ModelMeta(
         fp_options="float16",
     ),
     name="google/flan-t5-large",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="0613663d0d48ea86ba8cb3d7a44f0f65dc596a2a",
@@ -530,7 +531,6 @@ flant5_large = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
 )
 
 flant5_xl = ModelMeta(
@@ -539,6 +539,7 @@ flant5_xl = ModelMeta(
         fp_options="float16",
     ),
     name="google/flan-t5-xl",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="7d6315df2c2fb742f0f5b556879d730926ca9001",
@@ -576,7 +577,6 @@ flant5_xl = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
 )
 
 flant5_xxl = ModelMeta(
@@ -585,6 +585,7 @@ flant5_xxl = ModelMeta(
         fp_options="float16",
     ),
     name="google/flan-t5-xxl",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="ae7c9136adc7555eeccc78cdd960dfd60fb346ce",
@@ -622,7 +623,6 @@ flant5_xxl = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
 )
 
 
@@ -632,6 +632,7 @@ llama2_7b = ModelMeta(
         fp_options="float16",
     ),
     name="meta-llama/Llama-2-7b-hf",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="01c7f73d771dfac7d292323805ebc428287df4f9",
@@ -656,7 +657,6 @@ llama2_7b = ModelMeta(
       primaryClass={cs.CL},
       url={https://arxiv.org/abs/2307.09288},
     }""",
-    is_cross_encoder=True,
 )
 
 llama2_7b_chat = ModelMeta(
@@ -665,6 +665,7 @@ llama2_7b_chat = ModelMeta(
         fp_options="float16",
     ),
     name="meta-llama/Llama-2-7b-chat-hf",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="f5db02db724555f92da89c216ac04704f23d4590",
@@ -689,7 +690,6 @@ llama2_7b_chat = ModelMeta(
     use_instructions=None,
     training_datasets=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
 )
 
 mistral_7b = ModelMeta(
@@ -698,6 +698,7 @@ mistral_7b = ModelMeta(
         fp_options="float16",
     ),
     name="mistralai/Mistral-7B-Instruct-v0.2",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="3ad372fc79158a2148299e3318516c786aeded6c",
@@ -722,7 +723,6 @@ mistral_7b = ModelMeta(
       primaryClass={cs.CL},
       url={https://arxiv.org/abs/2310.06825},
     }""",
-    is_cross_encoder=True,
 )
 
 followir_7b = ModelMeta(
@@ -731,6 +731,7 @@ followir_7b = ModelMeta(
         fp_options="float16",
     ),
     name="jhu-clsp/FollowIR-7B",
+    model_type=["cross-encoder"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="4d25d437e38b510c01852070c0731e8f6e1875d1",
@@ -758,7 +759,6 @@ followir_7b = ModelMeta(
       primaryClass={cs.IR}
     }
     """,
-    is_cross_encoder=True,
 )
 
 
@@ -874,6 +874,7 @@ mt5_base_mmarco_v2 = ModelMeta(
         fp_options="float16",
     ),
     name="unicamp-dl/mt5-base-mmarco-v2",
+    model_type=["cross-encoder"],
     languages=mt5_languages,
     open_weights=True,
     revision="cc0a949b9f21efcaba45c8cabb998ad02ce8d4e7",
@@ -898,7 +899,6 @@ mt5_base_mmarco_v2 = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
 )
 
 mt5_13b_mmarco_100k = ModelMeta(
@@ -907,6 +907,7 @@ mt5_13b_mmarco_100k = ModelMeta(
         fp_options="float16",
     ),
     name="unicamp-dl/mt5-13b-mmarco-100k",
+    model_type=["cross-encoder"],
     languages=mt5_languages,
     open_weights=True,
     revision="e1a4317e102a525ea9e16745ad21394a4f1bffbc",
@@ -922,5 +923,4 @@ mt5_13b_mmarco_100k = ModelMeta(
     use_instructions=None,
     training_datasets=None,
     framework=["PyTorch"],
-    is_cross_encoder=True,
 )

mteb/models/model_implementations/richinfoai_models.py

@@ -9,6 +9,7 @@ from .stella_models import stella_zh_datasets
 ritrieve_zh_v1 = ModelMeta(
     loader=SentenceTransformerEncoderWrapper,
     name="richinfoai/ritrieve_zh_v1",
+    model_type=["dense"],
     languages=["zho-Hans"],
     open_weights=True,
     revision="f8d5a707656c55705027678e311f9202c8ced12c",

mteb/models/model_implementations/ru_sentence_models.py

@@ -238,6 +238,7 @@ GIGA_task_prompts = {
 rubert_tiny = ModelMeta(
     loader=sentence_transformers_loader,
     name="cointegrated/rubert-tiny",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="5441c5ea8026d4f6d7505ec004845409f1259fb1",
@@ -263,6 +264,7 @@ rubert_tiny = ModelMeta(
 rubert_tiny2 = ModelMeta(
     loader=sentence_transformers_loader,
     name="cointegrated/rubert-tiny2",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="dad72b8f77c5eef6995dd3e4691b758ba56b90c3",
@@ -289,6 +291,7 @@ rubert_tiny2 = ModelMeta(
 sbert_large_nlu_ru = ModelMeta(
     loader=sentence_transformers_loader,
     name="ai-forever/sbert_large_nlu_ru",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="af977d5dfa46a3635e29bf0ef383f2df2a08d47a",
@@ -314,6 +317,7 @@ sbert_large_nlu_ru = ModelMeta(
 sbert_large_mt_nlu_ru = ModelMeta(
     loader=sentence_transformers_loader,
     name="ai-forever/sbert_large_mt_nlu_ru",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="05300876c2b83f46d3ddd422a7f17e45cf633bb0",
@@ -341,6 +345,7 @@ user_base_ru = ModelMeta(
         model_prompts={"query": "query: ", "document": "passage: "},
     ),
     name="deepvk/USER-base",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="436a489a2087d61aa670b3496a9915f84e46c861",
@@ -401,6 +406,7 @@ user_base_ru = ModelMeta(
 user_bge_m3 = ModelMeta(
     loader=sentence_transformers_loader,
     name="deepvk/USER-bge-m3",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="0cc6cfe48e260fb0474c753087a69369e88709ae",
@@ -451,6 +457,7 @@ user_bge_m3 = ModelMeta(
 deberta_v1_ru = ModelMeta(
     loader=sentence_transformers_loader,
     name="deepvk/deberta-v1-base",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="bdd30b0e19757e6940c92c7aff19e8fc0a60dff4",
@@ -481,6 +488,7 @@ deberta_v1_ru = ModelMeta(
 rubert_base_cased = ModelMeta(
     loader=sentence_transformers_loader,
     name="DeepPavlov/rubert-base-cased",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="4036cab694767a299f2b9e6492909664d9414229",
@@ -516,6 +524,7 @@ rubert_base_cased = ModelMeta(
 distilrubert_small_cased_conversational = ModelMeta(
     loader=sentence_transformers_loader,
     name="DeepPavlov/distilrubert-small-cased-conversational",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="e348066b4a7279b97138038299bddc6580a9169a",
@@ -550,6 +559,7 @@ distilrubert_small_cased_conversational = ModelMeta(
 rubert_base_cased_sentence = ModelMeta(
     loader=sentence_transformers_loader,
     name="DeepPavlov/rubert-base-cased-sentence",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="78b5122d6365337dd4114281b0d08cd1edbb3bc8",
@@ -574,6 +584,7 @@ rubert_base_cased_sentence = ModelMeta(
 labse_en_ru = ModelMeta(
     loader=sentence_transformers_loader,
     name="cointegrated/LaBSE-en-ru",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="cf0714e606d4af551e14ad69a7929cd6b0da7f7e",
@@ -601,6 +612,7 @@ turbo_models_datasets = set(
 rubert_tiny_turbo = ModelMeta(
     loader=sentence_transformers_loader,
     name="sergeyzh/rubert-tiny-turbo",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="8ce0cf757446ce9bb2d5f5a4ac8103c7a1049054",
@@ -623,6 +635,7 @@ rubert_tiny_turbo = ModelMeta(
 rubert_mini_frida = ModelMeta(
     loader=sentence_transformers_loader,
     name="sergeyzh/rubert-mini-frida",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="19b279b78afd945b5ccae78f63e284909814adc2",
@@ -650,6 +663,7 @@ rubert_mini_frida = ModelMeta(
 labse_ru_turbo = ModelMeta(
     loader=sentence_transformers_loader,
     name="sergeyzh/LaBSE-ru-turbo",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="1940b046c6b5e125df11722b899130329d0a46da",
@@ -698,6 +712,7 @@ rosberta_ru_en = ModelMeta(
         model_prompts=rosberta_prompts,
     ),
     name="ai-forever/ru-en-RoSBERTa",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="89fb1651989adbb1cfcfdedafd7d102951ad0555",
@@ -863,6 +878,7 @@ frida = ModelMeta(
         model_prompts=frida_prompts,
     ),
     name="ai-forever/FRIDA",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="7292217af9a9e6dbf07048f76b434ad1e2aa8b76",
@@ -896,6 +912,7 @@ giga_embeddings = ModelMeta(
         },
     ),
     name="ai-sage/Giga-Embeddings-instruct",
+    model_type=["dense"],
     languages=["eng-Latn", "rus-Cyrl"],
     open_weights=True,
     revision="0ad5b29bfecd806cecc9d66b927d828a736594dc",
@@ -927,6 +944,7 @@ berta_training_datasets = (
 berta = ModelMeta(
     loader=sentence_transformers_loader,
     name="sergeyzh/BERTA",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="914c8c8aed14042ed890fc2c662d5e9e66b2faa7",
@@ -999,6 +1017,7 @@ user2_small = ModelMeta(
         model_prompts=user2_prompts,
     ),
     name="deepvk/USER2-small",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="23f65b34cf7632032061f5cc66c14714e6d4cee4",
@@ -1031,6 +1050,7 @@ user2_base = ModelMeta(
         model_prompts=user2_prompts,
     ),
     name="deepvk/USER2-base",
+    model_type=["dense"],
     languages=["rus-Cyrl"],
     open_weights=True,
     revision="0942cf96909b6d52e61f79a01e2d30c7be640b27",

mteb/models/model_implementations/ruri_models.py

@@ -32,6 +32,7 @@ cl_nagoya_ruri_v3_30m = ModelMeta(
         model_prompts=RURI_V3_PROMPTS,
     ),
     name="cl-nagoya/ruri-v3-30m",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="24899e5de370b56d179604a007c0d727bf144504",
@@ -62,6 +63,7 @@ cl_nagoya_ruri_v3_70m = ModelMeta(
         model_prompts=RURI_V3_PROMPTS,
     ),
     name="cl-nagoya/ruri-v3-70m",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="07a8b0aba47d29d2ca21f89b915c1efe2c23d1cc",
@@ -90,6 +92,7 @@ cl_nagoya_ruri_v3_130m = ModelMeta(
         model_prompts=RURI_V3_PROMPTS,
     ),
     name="cl-nagoya/ruri-v3-130m",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="e3114c6ee10dbab8b4b235fbc6dcf9dd4d5ac1a6",
@@ -118,6 +121,7 @@ cl_nagoya_ruri_v3_310m = ModelMeta(
         model_prompts=RURI_V3_PROMPTS,
     ),
     name="cl-nagoya/ruri-v3-310m",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="18b60fb8c2b9df296fb4212bb7d23ef94e579cd3",
@@ -147,6 +151,7 @@ cl_nagoya_ruri_small_v2 = ModelMeta(
         trust_remote_code=True,
     ),
     name="cl-nagoya/ruri-small-v2",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="db18646e673b713cd0518a5bb0fefdce21e77cd9",
@@ -175,6 +180,7 @@ cl_nagoya_ruri_base_v2 = ModelMeta(
         model_prompts=RURI_V1_V2_PROMPTS,
     ),
     name="cl-nagoya/ruri-base-v2",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="8ce03882903668a01c83ca3b8111ac025a3bc734",
@@ -203,6 +209,7 @@ cl_nagoya_ruri_large_v2 = ModelMeta(
         model_prompts=RURI_V1_V2_PROMPTS,
     ),
     name="cl-nagoya/ruri-large-v2",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="42898ef34a5574977380ebf0dfd28cbfbd36438b",
@@ -232,6 +239,7 @@ cl_nagoya_ruri_small_v1 = ModelMeta(
         trust_remote_code=True,
     ),
     name="cl-nagoya/ruri-small",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="bc56ce90cd7a979f6eb199fc52dfe700bfd94bc3",
@@ -260,6 +268,7 @@ cl_nagoya_ruri_base_v1 = ModelMeta(
         model_prompts=RURI_V1_V2_PROMPTS,
     ),
     name="cl-nagoya/ruri-base",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="1ae40b8b6c78518a499425086bab8fc16c2e4b0e",
@@ -289,6 +298,7 @@ cl_nagoya_ruri_large_v1 = ModelMeta(
         model_prompts=RURI_V1_V2_PROMPTS,
     ),
     name="cl-nagoya/ruri-large",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="a011c39b13e8bc137ee13c6bc82191ece46c414c",

mteb/models/model_implementations/salesforce_models.py

@@ -46,6 +46,7 @@ SFR_Embedding_2_R = ModelMeta(
         normalized=True,
     ),
     name="Salesforce/SFR-Embedding-2_R",
+    model_type=["dense"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="91762139d94ed4371a9fa31db5551272e0b83818",
@@ -83,6 +84,7 @@ SFR_Embedding_Code_2B_R = ModelMeta(
         normalized=True,
     ),
     name="Salesforce/SFR-Embedding-Code-2B_R",
+    model_type=["dense"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="c73d8631a005876ed5abde34db514b1fb6566973",
@@ -120,6 +122,7 @@ SFR_Embedding_Mistral = ModelMeta(
         normalized=True,
     ),
     name="Salesforce/SFR-Embedding-Mistral",
+    model_type=["dense"],
     languages=["eng-Latn"],
     open_weights=True,
     revision="938c560d1c236aa563b2dbdf084f28ab28bccb11",

mteb/models/model_implementations/samilpwc_models.py

@@ -43,6 +43,7 @@ samilpwc_expr = ModelMeta(
         apply_instruction_to_passages=False,
     ),
     name="SamilPwC-AXNode-GenAI/PwC-Embedding_expr",
+    model_type=["dense"],
     languages=[
         "kor-Hang",
     ],

mteb/models/model_implementations/sarashina_embedding_models.py

@@ -118,6 +118,7 @@ sbintuitions_sarashina_embedding_v2_1b = ModelMeta(
         max_seq_length=8192,
     ),
     name="sbintuitions/sarashina-embedding-v2-1b",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="1f3408afaa7b617e3445d891310a9c26dd0c68a5",
@@ -143,6 +144,7 @@ sbintuitions_sarashina_embedding_v2_1b = ModelMeta(
 sbintuitions_sarashina_embedding_v1_1b = ModelMeta(
     loader=sentence_transformers_loader,
     name="sbintuitions/sarashina-embedding-v1-1b",
+    model_type=["dense"],
     languages=["jpn-Jpan"],
     open_weights=True,
     revision="d060fcd8984075071e7fad81baff035cbb3b6c7e",

mteb/models/model_implementations/searchmap_models.py

@@ -20,6 +20,7 @@ searchmap_preview = ModelMeta(
         "model_prompts": task_instructions,
     },
     name="VPLabs/SearchMap_Preview",
+    model_type=["dense"],
     revision="69de17ef48278ed08ba1a4e65ead8179912b696e",
     languages=["eng-Latn"],
     open_weights=True,

mteb/models/model_implementations/seed_1_6_embedding_models.py

@@ -413,6 +413,7 @@ TASK_NAME_TO_INSTRUCTION = {
 
 seed_embedding = ModelMeta(
     name="Bytedance/Seed1.6-embedding",
+    model_type=["dense"],
     revision="1",
     release_date="2025-06-18",
     languages=[

mteb/models/model_implementations/seed_models.py

@@ -236,6 +236,7 @@ DEFAULT_INSTRUCTION = "Instruct: Given a web search query, retrieve relevant pas
 
 seed_embedding = ModelMeta(
     name="ByteDance-Seed/Seed1.5-Embedding",
+    model_type=["dense"],
     revision="4",
     release_date="2025-04-25",
     languages=[