mteb 2.5.1__py3-none-any.whl → 2.5.3__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- mteb/abstasks/abstask.py +6 -6
- mteb/abstasks/aggregated_task.py +4 -10
- mteb/abstasks/clustering_legacy.py +3 -2
- mteb/abstasks/task_metadata.py +2 -3
- mteb/cache.py +7 -4
- mteb/cli/build_cli.py +10 -5
- mteb/cli/generate_model_card.py +4 -3
- mteb/deprecated_evaluator.py +4 -3
- mteb/evaluate.py +4 -1
- mteb/get_tasks.py +4 -3
- mteb/leaderboard/app.py +70 -3
- mteb/models/abs_encoder.py +5 -3
- mteb/models/cache_wrappers/cache_backends/faiss_cache.py +4 -1
- mteb/models/cache_wrappers/cache_backends/numpy_cache.py +13 -12
- mteb/models/model_implementations/align_models.py +1 -0
- mteb/models/model_implementations/amazon_models.py +1 -0
- mteb/models/model_implementations/andersborges.py +2 -0
- mteb/models/model_implementations/ara_models.py +1 -0
- mteb/models/model_implementations/arctic_models.py +8 -0
- mteb/models/model_implementations/b1ade_models.py +1 -0
- mteb/models/model_implementations/bedrock_models.py +4 -0
- mteb/models/model_implementations/bge_models.py +17 -0
- mteb/models/model_implementations/bica_model.py +1 -0
- mteb/models/model_implementations/blip2_models.py +2 -0
- mteb/models/model_implementations/blip_models.py +8 -0
- mteb/models/model_implementations/bm25.py +1 -0
- mteb/models/model_implementations/bmretriever_models.py +4 -0
- mteb/models/model_implementations/cadet_models.py +1 -0
- mteb/models/model_implementations/cde_models.py +2 -0
- mteb/models/model_implementations/clip_models.py +3 -0
- mteb/models/model_implementations/clips_models.py +3 -0
- mteb/models/model_implementations/codefuse_models.py +3 -0
- mteb/models/model_implementations/codesage_models.py +3 -0
- mteb/models/model_implementations/cohere_models.py +4 -0
- mteb/models/model_implementations/cohere_v.py +5 -0
- mteb/models/model_implementations/colpali_models.py +3 -0
- mteb/models/model_implementations/colqwen_models.py +9 -0
- mteb/models/model_implementations/colsmol_models.py +2 -0
- mteb/models/model_implementations/conan_models.py +1 -0
- mteb/models/model_implementations/dino_models.py +19 -0
- mteb/models/model_implementations/e5_instruct.py +4 -0
- mteb/models/model_implementations/e5_models.py +9 -0
- mteb/models/model_implementations/e5_v.py +1 -0
- mteb/models/model_implementations/eagerworks_models.py +1 -0
- mteb/models/model_implementations/emillykkejensen_models.py +3 -0
- mteb/models/model_implementations/en_code_retriever.py +1 -0
- mteb/models/model_implementations/euler_models.py +1 -0
- mteb/models/model_implementations/evaclip_models.py +4 -0
- mteb/models/model_implementations/fa_models.py +8 -0
- mteb/models/model_implementations/facebookai.py +2 -0
- mteb/models/model_implementations/geogpt_models.py +1 -0
- mteb/models/model_implementations/gme_v_models.py +6 -3
- mteb/models/model_implementations/google_models.py +5 -0
- mteb/models/model_implementations/granite_vision_embedding_models.py +1 -0
- mteb/models/model_implementations/gritlm_models.py +2 -0
- mteb/models/model_implementations/gte_models.py +9 -0
- mteb/models/model_implementations/hinvec_models.py +1 -0
- mteb/models/model_implementations/human.py +1 -0
- mteb/models/model_implementations/ibm_granite_models.py +6 -0
- mteb/models/model_implementations/inf_models.py +2 -0
- mteb/models/model_implementations/jasper_models.py +2 -0
- mteb/models/model_implementations/jina_clip.py +1 -0
- mteb/models/model_implementations/jina_models.py +7 -1
- mteb/models/model_implementations/kalm_models.py +6 -0
- mteb/models/model_implementations/kblab.py +1 -0
- mteb/models/model_implementations/kennethenevoldsen_models.py +2 -0
- mteb/models/model_implementations/kfst.py +1 -0
- mteb/models/model_implementations/kowshik24_models.py +1 -0
- mteb/models/model_implementations/lens_models.py +2 -0
- mteb/models/model_implementations/lgai_embedding_models.py +1 -0
- mteb/models/model_implementations/linq_models.py +1 -0
- mteb/models/model_implementations/listconranker.py +1 -1
- mteb/models/model_implementations/llm2clip_models.py +3 -0
- mteb/models/model_implementations/llm2vec_models.py +8 -0
- mteb/models/model_implementations/mcinext_models.py +7 -1
- mteb/models/model_implementations/mdbr_models.py +2 -0
- mteb/models/model_implementations/misc_models.py +63 -0
- mteb/models/model_implementations/mme5_models.py +1 -0
- mteb/models/model_implementations/moco_models.py +2 -0
- mteb/models/model_implementations/model2vec_models.py +13 -0
- mteb/models/model_implementations/moka_models.py +3 -0
- mteb/models/model_implementations/mxbai_models.py +3 -0
- mteb/models/model_implementations/nbailab.py +3 -0
- mteb/models/model_implementations/no_instruct_sentence_models.py +1 -0
- mteb/models/model_implementations/nomic_models.py +6 -0
- mteb/models/model_implementations/nomic_models_vision.py +1 -0
- mteb/models/model_implementations/nvidia_llama_nemoretriever_colemb.py +2 -0
- mteb/models/model_implementations/nvidia_models.py +3 -0
- mteb/models/model_implementations/octen_models.py +195 -0
- mteb/models/model_implementations/openai_models.py +5 -0
- mteb/models/model_implementations/openclip_models.py +8 -0
- mteb/models/model_implementations/opensearch_neural_sparse_models.py +5 -0
- mteb/models/model_implementations/ops_moa_models.py +2 -0
- mteb/models/model_implementations/pawan_models.py +1 -0
- mteb/models/model_implementations/piccolo_models.py +2 -0
- mteb/models/model_implementations/promptriever_models.py +4 -0
- mteb/models/model_implementations/pylate_models.py +3 -0
- mteb/models/model_implementations/qodo_models.py +2 -0
- mteb/models/model_implementations/qtack_models.py +1 -0
- mteb/models/model_implementations/qwen3_models.py +3 -0
- mteb/models/model_implementations/qzhou_models.py +2 -0
- mteb/models/model_implementations/random_baseline.py +2 -1
- mteb/models/model_implementations/rasgaard_models.py +1 -0
- mteb/models/model_implementations/reasonir_model.py +1 -0
- mteb/models/model_implementations/repllama_models.py +2 -0
- mteb/models/model_implementations/rerankers_custom.py +3 -3
- mteb/models/model_implementations/rerankers_monot5_based.py +14 -14
- mteb/models/model_implementations/richinfoai_models.py +1 -0
- mteb/models/model_implementations/ru_sentence_models.py +20 -0
- mteb/models/model_implementations/ruri_models.py +10 -0
- mteb/models/model_implementations/salesforce_models.py +3 -0
- mteb/models/model_implementations/samilpwc_models.py +1 -0
- mteb/models/model_implementations/sarashina_embedding_models.py +2 -0
- mteb/models/model_implementations/searchmap_models.py +1 -0
- mteb/models/model_implementations/seed_1_6_embedding_models.py +1 -0
- mteb/models/model_implementations/seed_models.py +1 -0
- mteb/models/model_implementations/sentence_transformers_models.py +18 -0
- mteb/models/model_implementations/shuu_model.py +32 -31
- mteb/models/model_implementations/siglip_models.py +10 -0
- mteb/models/model_implementations/sonar_models.py +1 -0
- mteb/models/model_implementations/spartan8806_atles_champion.py +1 -0
- mteb/models/model_implementations/stella_models.py +6 -0
- mteb/models/model_implementations/tarka_models.py +2 -0
- mteb/models/model_implementations/ua_sentence_models.py +1 -0
- mteb/models/model_implementations/uae_models.py +1 -0
- mteb/models/model_implementations/vdr_models.py +1 -0
- mteb/models/model_implementations/vi_vn_models.py +6 -0
- mteb/models/model_implementations/vista_models.py +2 -0
- mteb/models/model_implementations/vlm2vec_models.py +2 -0
- mteb/models/model_implementations/voyage_models.py +15 -0
- mteb/models/model_implementations/voyage_v.py +1 -0
- mteb/models/model_implementations/xyz_models.py +1 -0
- mteb/models/model_implementations/youtu_models.py +1 -0
- mteb/models/model_implementations/yuan_models.py +1 -0
- mteb/models/model_implementations/yuan_models_en.py +1 -0
- mteb/models/model_meta.py +49 -4
- mteb/models/search_encoder_index/search_indexes/faiss_search_index.py +4 -1
- mteb/models/search_wrappers.py +4 -2
- mteb/models/sentence_transformer_wrapper.py +10 -10
- mteb/results/benchmark_results.py +67 -43
- mteb/results/model_result.py +3 -1
- mteb/results/task_result.py +22 -17
- {mteb-2.5.1.dist-info → mteb-2.5.3.dist-info}/METADATA +1 -1
- {mteb-2.5.1.dist-info → mteb-2.5.3.dist-info}/RECORD +148 -147
- {mteb-2.5.1.dist-info → mteb-2.5.3.dist-info}/WHEEL +0 -0
- {mteb-2.5.1.dist-info → mteb-2.5.3.dist-info}/entry_points.txt +0 -0
- {mteb-2.5.1.dist-info → mteb-2.5.3.dist-info}/licenses/LICENSE +0 -0
- {mteb-2.5.1.dist-info → mteb-2.5.3.dist-info}/top_level.txt +0 -0
|
@@ -50,6 +50,7 @@ inf_retriever_v1 = ModelMeta(
|
|
|
50
50
|
trust_remote_code=True,
|
|
51
51
|
),
|
|
52
52
|
name="infly/inf-retriever-v1",
|
|
53
|
+
model_type=["dense"],
|
|
53
54
|
languages=["eng-Latn", "zho-Hans"],
|
|
54
55
|
open_weights=True,
|
|
55
56
|
revision="cb70ca7c31dfa866b2eff2dad229c144d8ddfd91",
|
|
@@ -76,6 +77,7 @@ inf_retriever_v1_1_5b = ModelMeta(
|
|
|
76
77
|
trust_remote_code=True,
|
|
77
78
|
),
|
|
78
79
|
name="infly/inf-retriever-v1-1.5b",
|
|
80
|
+
model_type=["dense"],
|
|
79
81
|
languages=["eng-Latn", "zho-Hans"],
|
|
80
82
|
open_weights=True,
|
|
81
83
|
revision="c9c05c2dd50707a486966ba81703021ae2094a06",
|
|
@@ -286,6 +286,7 @@ jasper_en_v1 = ModelMeta(
|
|
|
286
286
|
instruction_template="Instruct: {instruction}\nQuery: ",
|
|
287
287
|
),
|
|
288
288
|
name="NovaSearch/jasper_en_vision_language_v1",
|
|
289
|
+
model_type=["dense"],
|
|
289
290
|
languages=["eng-Latn"],
|
|
290
291
|
open_weights=True,
|
|
291
292
|
revision="d6330ce98f8a0d741e781df845904c9484f00efa",
|
|
@@ -332,6 +333,7 @@ Jasper_Token_Compression_600M = ModelMeta(
|
|
|
332
333
|
loader=InstructSentenceTransformerModel,
|
|
333
334
|
loader_kwargs=jasper_token_compression_600m_loader_kwargs,
|
|
334
335
|
name="infgrad/Jasper-Token-Compression-600M",
|
|
336
|
+
model_type=["dense"],
|
|
335
337
|
languages=["eng-Latn", "zho-Hans"],
|
|
336
338
|
open_weights=True,
|
|
337
339
|
revision="06a100f753a5a96d9e583b3af79c6fcdfacc4719",
|
|
@@ -123,6 +123,7 @@ class JinaCLIPModel(AbsEncoder):
|
|
|
123
123
|
jina_clip_v1 = ModelMeta(
|
|
124
124
|
loader=JinaCLIPModel, # type: ignore
|
|
125
125
|
name="jinaai/jina-clip-v1",
|
|
126
|
+
model_type=["dense"],
|
|
126
127
|
languages=["eng-Latn"],
|
|
127
128
|
revision="06150c7c382d7a4faedc7d5a0d8cdb59308968f4",
|
|
128
129
|
release_date="2024-05-30",
|
|
@@ -720,6 +720,7 @@ jina_reranker_v3 = ModelMeta(
|
|
|
720
720
|
trust_remote_code=True,
|
|
721
721
|
),
|
|
722
722
|
name="jinaai/jina-reranker-v3",
|
|
723
|
+
model_type=["cross-encoder"],
|
|
723
724
|
languages=multilingual_langs,
|
|
724
725
|
open_weights=True,
|
|
725
726
|
revision="050e171c4f75dfec5b648ed8470a2475e5a30f30",
|
|
@@ -734,7 +735,6 @@ jina_reranker_v3 = ModelMeta(
|
|
|
734
735
|
framework=["PyTorch"],
|
|
735
736
|
use_instructions=None,
|
|
736
737
|
reference="https://huggingface.co/jinaai/jina-reranker-v3",
|
|
737
|
-
is_cross_encoder=True,
|
|
738
738
|
public_training_code=None,
|
|
739
739
|
public_training_data=None,
|
|
740
740
|
training_datasets=JINARerankerV3_TRAINING_DATA,
|
|
@@ -763,6 +763,7 @@ jina_embeddings_v4 = ModelMeta(
|
|
|
763
763
|
},
|
|
764
764
|
),
|
|
765
765
|
name="jinaai/jina-embeddings-v4",
|
|
766
|
+
model_type=["dense"],
|
|
766
767
|
languages=XLMR_LANGUAGES,
|
|
767
768
|
open_weights=True,
|
|
768
769
|
revision="4a58ca57710c49f51896e4bc820e202fbf64904b",
|
|
@@ -811,6 +812,7 @@ jina_embeddings_v3 = ModelMeta(
|
|
|
811
812
|
},
|
|
812
813
|
),
|
|
813
814
|
name="jinaai/jina-embeddings-v3",
|
|
815
|
+
model_type=["dense"],
|
|
814
816
|
languages=XLMR_LANGUAGES,
|
|
815
817
|
open_weights=True,
|
|
816
818
|
revision="215a6e121fa0183376388ac6b1ae230326bfeaed",
|
|
@@ -864,6 +866,7 @@ jina_embeddings_v2_base_en = ModelMeta(
|
|
|
864
866
|
trust_remote_code=True,
|
|
865
867
|
),
|
|
866
868
|
name="jinaai/jina-embeddings-v2-base-en",
|
|
869
|
+
model_type=["dense"],
|
|
867
870
|
languages=["eng-Latn"],
|
|
868
871
|
open_weights=True,
|
|
869
872
|
revision="6e85f575bc273f1fd840a658067d0157933c83f0",
|
|
@@ -927,6 +930,7 @@ jina_embeddings_v2_small_en = ModelMeta(
|
|
|
927
930
|
trust_remote_code=True,
|
|
928
931
|
),
|
|
929
932
|
name="jinaai/jina-embeddings-v2-small-en",
|
|
933
|
+
model_type=["dense"],
|
|
930
934
|
languages=["eng-Latn"],
|
|
931
935
|
open_weights=True,
|
|
932
936
|
revision="44e7d1d6caec8c883c2d4b207588504d519788d0",
|
|
@@ -987,6 +991,7 @@ jina_embeddings_v2_small_en = ModelMeta(
|
|
|
987
991
|
jina_embedding_b_en_v1 = ModelMeta(
|
|
988
992
|
loader=SentenceTransformerEncoderWrapper,
|
|
989
993
|
name="jinaai/jina-embedding-b-en-v1",
|
|
994
|
+
model_type=["dense"],
|
|
990
995
|
languages=["eng-Latn"],
|
|
991
996
|
open_weights=True,
|
|
992
997
|
revision="32aa658e5ceb90793454d22a57d8e3a14e699516",
|
|
@@ -1043,6 +1048,7 @@ jina_embedding_b_en_v1 = ModelMeta(
|
|
|
1043
1048
|
jina_embedding_s_en_v1 = ModelMeta(
|
|
1044
1049
|
loader=SentenceTransformerEncoderWrapper,
|
|
1045
1050
|
name="jinaai/jina-embedding-s-en-v1",
|
|
1051
|
+
model_type=["dense"],
|
|
1046
1052
|
languages=["eng-Latn"],
|
|
1047
1053
|
open_weights=True,
|
|
1048
1054
|
revision="5ac6cd473e2324c6d5f9e558a6a9f65abb57143e",
|
|
@@ -769,6 +769,7 @@ HIT_TMG__KaLM_embedding_multilingual_mini_instruct_v1 = ModelMeta(
|
|
|
769
769
|
prompts_dict=KaLM_task_prompts,
|
|
770
770
|
),
|
|
771
771
|
name="HIT-TMG/KaLM-embedding-multilingual-mini-instruct-v1",
|
|
772
|
+
model_type=["dense"],
|
|
772
773
|
revision="45e42c89990c40aca042659133fc8b13c28634b5",
|
|
773
774
|
release_date="2024-10-23",
|
|
774
775
|
languages=["eng-Latn", "zho-Hans"],
|
|
@@ -793,6 +794,7 @@ HIT_TMG__KaLM_embedding_multilingual_mini_instruct_v1 = ModelMeta(
|
|
|
793
794
|
HIT_TMG__KaLM_embedding_multilingual_mini_v1 = ModelMeta(
|
|
794
795
|
loader=sentence_transformers_loader,
|
|
795
796
|
name="HIT-TMG/KaLM-embedding-multilingual-mini-v1",
|
|
797
|
+
model_type=["dense"],
|
|
796
798
|
revision="8a82a0cd2b322b91723e252486f7cce6fd8ac9d3",
|
|
797
799
|
release_date="2024-08-27",
|
|
798
800
|
languages=["eng-Latn", "zho-Hans"],
|
|
@@ -823,6 +825,7 @@ HIT_TMG__KaLM_embedding_multilingual_mini_instruct_v1_5 = ModelMeta(
|
|
|
823
825
|
prompts_dict=KaLM_task_prompts,
|
|
824
826
|
),
|
|
825
827
|
name="HIT-TMG/KaLM-embedding-multilingual-mini-instruct-v1.5",
|
|
828
|
+
model_type=["dense"],
|
|
826
829
|
revision="fcff2f8a54e4cd96b7766fef1ee960a43d42bb3c",
|
|
827
830
|
release_date="2024-12-26",
|
|
828
831
|
languages=["eng-Latn", "zho-Hans"],
|
|
@@ -853,6 +856,7 @@ HIT_TMG__KaLM_embedding_multilingual_mini_instruct_v2 = ModelMeta(
|
|
|
853
856
|
prompts_dict=KaLM_v2_task_prompts,
|
|
854
857
|
),
|
|
855
858
|
name="HIT-TMG/KaLM-embedding-multilingual-mini-instruct-v2",
|
|
859
|
+
model_type=["dense"],
|
|
856
860
|
revision="d2a21c232dc712ae8230af56d1027cf21b7864bf",
|
|
857
861
|
release_date="2025-06-25",
|
|
858
862
|
languages=["eng-Latn", "zho-Hans"],
|
|
@@ -883,6 +887,7 @@ KaLM_Embedding_KaLM_embedding_multilingual_mini_instruct_v2_5 = ModelMeta(
|
|
|
883
887
|
prompts_dict=KaLM_v2_task_prompts,
|
|
884
888
|
),
|
|
885
889
|
name="KaLM-Embedding/KaLM-embedding-multilingual-mini-instruct-v2.5",
|
|
890
|
+
model_type=["dense"],
|
|
886
891
|
revision="6a4cfc1084cb459ebd4729b53a8656a61448c720",
|
|
887
892
|
release_date="2025-09-30",
|
|
888
893
|
languages=["eng-Latn", "zho-Hans"],
|
|
@@ -931,6 +936,7 @@ KaLM_Embedding_gemma_3_12b_2511 = ModelMeta(
|
|
|
931
936
|
prompts_dict=KaLM_Embedding_gemma_3_12b_task_prompts,
|
|
932
937
|
),
|
|
933
938
|
name="tencent/KaLM-Embedding-Gemma3-12B-2511",
|
|
939
|
+
model_type=["dense"],
|
|
934
940
|
revision="edf22f4753f58b05e3f5495818d31f12db63056d",
|
|
935
941
|
languages=None,
|
|
936
942
|
open_weights=True,
|
|
@@ -4,6 +4,7 @@ from mteb.models.model_meta import ModelMeta, ScoringFunction
|
|
|
4
4
|
sbert_swedish = ModelMeta(
|
|
5
5
|
loader=sentence_transformers_loader, # type: ignore[arg-type]
|
|
6
6
|
name="KBLab/sentence-bert-swedish-cased",
|
|
7
|
+
model_type=["dense"],
|
|
7
8
|
languages=["swe-Latn"],
|
|
8
9
|
open_weights=True,
|
|
9
10
|
revision="6b5e83cd29c03729cfdc33d13b1423399b0efb5c",
|
|
@@ -6,6 +6,7 @@ from mteb.models.sentence_transformer_wrapper import (
|
|
|
6
6
|
dfm_enc_large = ModelMeta(
|
|
7
7
|
loader=sentence_transformers_loader, # type: ignore
|
|
8
8
|
name="KennethEnevoldsen/dfm-sentence-encoder-large",
|
|
9
|
+
model_type=["dense"],
|
|
9
10
|
languages=["dan-Latn"],
|
|
10
11
|
open_weights=True,
|
|
11
12
|
revision="132c53391e7a780dc6a2f9a03724d0158fe7122c",
|
|
@@ -40,6 +41,7 @@ dfm_enc_large = ModelMeta(
|
|
|
40
41
|
dfm_enc_med = ModelMeta(
|
|
41
42
|
loader=sentence_transformers_loader, # type: ignore
|
|
42
43
|
name="KennethEnevoldsen/dfm-sentence-encoder-medium",
|
|
44
|
+
model_type=["dense"],
|
|
43
45
|
languages=["dan-Latn"],
|
|
44
46
|
open_weights=True,
|
|
45
47
|
revision="701bce95d499fa97610d57e8823c54fd1fb79930",
|
|
@@ -4,6 +4,7 @@ from mteb.models.model_meta import ModelMeta, ScoringFunction
|
|
|
4
4
|
xlmr_scandi = ModelMeta(
|
|
5
5
|
loader=sentence_transformers_loader, # type: ignore[arg-type]
|
|
6
6
|
name="KFST/XLMRoberta-en-da-sv-nb",
|
|
7
|
+
model_type=["dense"],
|
|
7
8
|
languages=["swe-Latn", "nob-Latn", "nno-Latn", "dan-Latn", "eng-Latn"],
|
|
8
9
|
open_weights=True,
|
|
9
10
|
revision="d40c10ca7b1e68b5a8372f2d112dac9eb3279df1",
|
|
@@ -3,6 +3,7 @@ from mteb.models import ModelMeta, sentence_transformers_loader
|
|
|
3
3
|
kowshik24_bangla_embedding_model = ModelMeta(
|
|
4
4
|
loader=sentence_transformers_loader,
|
|
5
5
|
name="Kowshik24/bangla-sentence-transformer-ft-matryoshka-paraphrase-multilingual-mpnet-base-v2",
|
|
6
|
+
model_type=["dense"],
|
|
6
7
|
languages=["ben-Beng"], # Bengali using Bengali script
|
|
7
8
|
open_weights=True,
|
|
8
9
|
revision="6689c21e69be5950596bad084457cbaa138728d8",
|
|
@@ -12,6 +12,7 @@ LENS_CITATION = """@article{lei2025lens,
|
|
|
12
12
|
lens_d4000 = ModelMeta(
|
|
13
13
|
loader=None,
|
|
14
14
|
name="yibinlei/LENS-d4000",
|
|
15
|
+
model_type=["dense"],
|
|
15
16
|
languages=None,
|
|
16
17
|
open_weights=True,
|
|
17
18
|
revision="e473b33364e6c48a324796fd1411d3b93670c6fe",
|
|
@@ -34,6 +35,7 @@ lens_d4000 = ModelMeta(
|
|
|
34
35
|
lens_d8000 = ModelMeta(
|
|
35
36
|
loader=None,
|
|
36
37
|
name="yibinlei/LENS-d8000",
|
|
38
|
+
model_type=["dense"],
|
|
37
39
|
languages=None,
|
|
38
40
|
open_weights=True,
|
|
39
41
|
revision="a0b87bd91cb27b6f2f0b0fe22c28026da1d464ef",
|
|
@@ -112,6 +112,7 @@ listconranker = ModelMeta(
|
|
|
112
112
|
fp_options="float16",
|
|
113
113
|
),
|
|
114
114
|
name="ByteDance/ListConRanker",
|
|
115
|
+
model_type=["cross-encoder"],
|
|
115
116
|
languages=["zho-Hans"],
|
|
116
117
|
open_weights=True,
|
|
117
118
|
revision="95ae6a5f422a916bc36520f0f3e198e7d91520a0",
|
|
@@ -128,6 +129,5 @@ listconranker = ModelMeta(
|
|
|
128
129
|
use_instructions=False,
|
|
129
130
|
public_training_code=None,
|
|
130
131
|
public_training_data=None,
|
|
131
|
-
is_cross_encoder=True,
|
|
132
132
|
citation=LISTCONRANKER_CITATION,
|
|
133
133
|
)
|
|
@@ -183,6 +183,7 @@ llm2clip_training_sets = set(
|
|
|
183
183
|
llm2clip_openai_l_14_336 = ModelMeta(
|
|
184
184
|
loader=llm2clip_loader, # type: ignore
|
|
185
185
|
name="microsoft/LLM2CLIP-Openai-L-14-336",
|
|
186
|
+
model_type=["dense"],
|
|
186
187
|
languages=["eng-Latn"],
|
|
187
188
|
revision="92512331f393a003c3d98404677f991c188162c9",
|
|
188
189
|
release_date="2024-11-07",
|
|
@@ -207,6 +208,7 @@ llm2clip_openai_l_14_336 = ModelMeta(
|
|
|
207
208
|
llm2clip_openai_l_14_224 = ModelMeta(
|
|
208
209
|
loader=llm2clip_loader, # type: ignore
|
|
209
210
|
name="microsoft/LLM2CLIP-Openai-L-14-224",
|
|
211
|
+
model_type=["dense"],
|
|
210
212
|
languages=["eng-Latn"],
|
|
211
213
|
revision="6b8a11a94ff380fa220dfefe73ac9293d2677575",
|
|
212
214
|
release_date="2024-11-07",
|
|
@@ -230,6 +232,7 @@ llm2clip_openai_l_14_224 = ModelMeta(
|
|
|
230
232
|
llm2clip_openai_b_16 = ModelMeta(
|
|
231
233
|
loader=llm2clip_loader, # type: ignore
|
|
232
234
|
name="microsoft/LLM2CLIP-Openai-B-16",
|
|
235
|
+
model_type=["dense"],
|
|
233
236
|
languages=["eng-Latn"],
|
|
234
237
|
revision="ecfb347eb3dcfeb2fbc2a2eae7de6ac5a001aaf8",
|
|
235
238
|
release_date="2024-11-07",
|
|
@@ -132,6 +132,7 @@ llm2vec_llama3_8b_supervised = ModelMeta(
|
|
|
132
132
|
torch_dtype=torch.bfloat16,
|
|
133
133
|
),
|
|
134
134
|
name="McGill-NLP/LLM2Vec-Meta-Llama-3-8B-Instruct-mntp-supervised",
|
|
135
|
+
model_type=["dense"],
|
|
135
136
|
languages=["eng-Latn"],
|
|
136
137
|
open_weights=True,
|
|
137
138
|
revision="baa8ebf04a1c2500e61288e7dad65e8ae42601a7",
|
|
@@ -161,6 +162,7 @@ llm2vec_llama3_8b_unsupervised = ModelMeta(
|
|
|
161
162
|
torch_dtype=torch.bfloat16,
|
|
162
163
|
),
|
|
163
164
|
name="McGill-NLP/LLM2Vec-Meta-Llama-3-8B-Instruct-mntp-unsup-simcse",
|
|
165
|
+
model_type=["dense"],
|
|
164
166
|
languages=["eng-Latn"],
|
|
165
167
|
open_weights=True,
|
|
166
168
|
revision="1cb7b735326d13a8541db8f57f35da5373f5e9c6",
|
|
@@ -189,6 +191,7 @@ llm2vec_mistral7b_supervised = ModelMeta(
|
|
|
189
191
|
torch_dtype=torch.bfloat16,
|
|
190
192
|
),
|
|
191
193
|
name="McGill-NLP/LLM2Vec-Mistral-7B-Instruct-v2-mntp-supervised",
|
|
194
|
+
model_type=["dense"],
|
|
192
195
|
languages=["eng-Latn"],
|
|
193
196
|
open_weights=True,
|
|
194
197
|
revision="0ae69bdd5816105778b971c3138e8f8a18eaa3ae",
|
|
@@ -217,6 +220,7 @@ llm2vec_mistral7b_unsupervised = ModelMeta(
|
|
|
217
220
|
torch_dtype=torch.bfloat16,
|
|
218
221
|
),
|
|
219
222
|
name="McGill-NLP/LLM2Vec-Mistral-7B-Instruct-v2-mntp-unsup-simcse",
|
|
223
|
+
model_type=["dense"],
|
|
220
224
|
languages=["eng-Latn"],
|
|
221
225
|
open_weights=True,
|
|
222
226
|
revision="2c055a5d77126c0d3dc6cd8ffa30e2908f4f45f8",
|
|
@@ -245,6 +249,7 @@ llm2vec_llama2_7b_supervised = ModelMeta(
|
|
|
245
249
|
torch_dtype=torch.bfloat16,
|
|
246
250
|
),
|
|
247
251
|
name="McGill-NLP/LLM2Vec-Llama-2-7b-chat-hf-mntp-supervised",
|
|
252
|
+
model_type=["dense"],
|
|
248
253
|
languages=["eng-Latn"],
|
|
249
254
|
open_weights=True,
|
|
250
255
|
revision="2c055a5d77126c0d3dc6cd8ffa30e2908f4f45f8",
|
|
@@ -273,6 +278,7 @@ llm2vec_llama2_7b_unsupervised = ModelMeta(
|
|
|
273
278
|
torch_dtype=torch.bfloat16,
|
|
274
279
|
),
|
|
275
280
|
name="McGill-NLP/LLM2Vec-Llama-2-7b-chat-hf-mntp-unsup-simcse",
|
|
281
|
+
model_type=["dense"],
|
|
276
282
|
languages=["eng-Latn"],
|
|
277
283
|
open_weights=True,
|
|
278
284
|
revision="a76944871d169ebe7c97eb921764cd063afed785",
|
|
@@ -301,6 +307,7 @@ llm2vec_sheared_llama_supervised = ModelMeta(
|
|
|
301
307
|
torch_dtype=torch.bfloat16,
|
|
302
308
|
),
|
|
303
309
|
name="McGill-NLP/LLM2Vec-Sheared-LLaMA-mntp-supervised",
|
|
310
|
+
model_type=["dense"],
|
|
304
311
|
languages=["eng-Latn"],
|
|
305
312
|
open_weights=True,
|
|
306
313
|
revision="a5943d406c6b016fef3f07906aac183cf1a0b47d",
|
|
@@ -329,6 +336,7 @@ llm2vec_sheared_llama_unsupervised = ModelMeta(
|
|
|
329
336
|
torch_dtype=torch.bfloat16,
|
|
330
337
|
),
|
|
331
338
|
name="McGill-NLP/LLM2Vec-Sheared-LLaMA-mntp-unsup-simcse",
|
|
339
|
+
model_type=["dense"],
|
|
332
340
|
languages=["eng-Latn"],
|
|
333
341
|
open_weights=True,
|
|
334
342
|
revision="a5943d406c6b016fef3f07906aac183cf1a0b47d",
|
|
@@ -1,6 +1,7 @@
|
|
|
1
1
|
import logging
|
|
2
2
|
import os
|
|
3
3
|
import time
|
|
4
|
+
import warnings
|
|
4
5
|
from typing import Any
|
|
5
6
|
|
|
6
7
|
import numpy as np
|
|
@@ -246,7 +247,9 @@ class HakimModelWrapper(AbsEncoder):
|
|
|
246
247
|
task_prompt, task_id = DATASET_TASKS.get(task_name, (None, None))
|
|
247
248
|
|
|
248
249
|
if not task_prompt:
|
|
249
|
-
|
|
250
|
+
msg = f"Unknown dataset: {task_name}, no preprocessing applied."
|
|
251
|
+
logger.warning(msg)
|
|
252
|
+
warnings.warn(msg)
|
|
250
253
|
return sample
|
|
251
254
|
|
|
252
255
|
task_prompt = f"مسئله : {task_prompt}"
|
|
@@ -344,6 +347,7 @@ hakim = ModelMeta(
|
|
|
344
347
|
loader=HakimModelWrapper,
|
|
345
348
|
loader_kwargs=dict(
|
|
346
349
|
api_model_name="hakim",
|
|
350
|
+
model_type=["dense"],
|
|
347
351
|
),
|
|
348
352
|
name="MCINext/Hakim",
|
|
349
353
|
languages=["fas-Arab"],
|
|
@@ -411,6 +415,7 @@ hakim_small = ModelMeta(
|
|
|
411
415
|
loader=HakimModelWrapper,
|
|
412
416
|
loader_kwargs=dict(
|
|
413
417
|
api_model_name="hakim-small",
|
|
418
|
+
model_type=["dense"],
|
|
414
419
|
),
|
|
415
420
|
name="MCINext/Hakim-small",
|
|
416
421
|
languages=["fas-Arab"],
|
|
@@ -477,6 +482,7 @@ hakim_unsup = ModelMeta(
|
|
|
477
482
|
loader=HakimModelWrapper,
|
|
478
483
|
loader_kwargs=dict(
|
|
479
484
|
api_model_name="hakim-unsup",
|
|
485
|
+
model_type=["dense"],
|
|
480
486
|
),
|
|
481
487
|
name="MCINext/Hakim-unsup",
|
|
482
488
|
languages=["fas-Arab"],
|
|
@@ -30,6 +30,7 @@ mdbr_leaf_ir = ModelMeta(
|
|
|
30
30
|
model_prompts=model_prompts,
|
|
31
31
|
),
|
|
32
32
|
name="MongoDB/mdbr-leaf-ir",
|
|
33
|
+
model_type=["dense"],
|
|
33
34
|
revision="2e46f5aac796e621d51f678c306a66ede4712ecb",
|
|
34
35
|
release_date="2025-08-27",
|
|
35
36
|
languages=["eng-Latn"],
|
|
@@ -57,6 +58,7 @@ mdbr_leaf_mt = ModelMeta(
|
|
|
57
58
|
model_prompts=model_prompts,
|
|
58
59
|
),
|
|
59
60
|
name="MongoDB/mdbr-leaf-mt",
|
|
61
|
+
model_type=["dense"],
|
|
60
62
|
revision="66c47ba6d753efc208d54412b5af6c744a39a4df",
|
|
61
63
|
release_date="2025-08-27",
|
|
62
64
|
languages=["eng-Latn"],
|