mteb 2.5.2__py3-none-any.whl → 2.7.2__py3-none-any.whl

mteb/models/model_implementations/opensearch_neural_sparse_models.py

@@ -140,7 +140,7 @@ opensearch_neural_sparse_encoding_doc_v3_gte = ModelMeta(
     max_tokens=8192,
     reference="https://huggingface.co/opensearch-project/opensearch-neural-sparse-encoding-doc-v3-gte",
     similarity_fn_name="dot",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     public_training_code="https://github.com/zhichao-aws/opensearch-sparse-model-tuning-sample",
     public_training_data=True,
     use_instructions=True,
@@ -166,7 +166,7 @@ opensearch_neural_sparse_encoding_doc_v3_distill = ModelMeta(
     max_tokens=512,
     reference="https://huggingface.co/opensearch-project/opensearch-neural-sparse-encoding-doc-v3-distill",
     similarity_fn_name="dot",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     public_training_code="https://github.com/zhichao-aws/opensearch-sparse-model-tuning-sample",
     public_training_data=True,
     use_instructions=True,
@@ -188,7 +188,7 @@ opensearch_neural_sparse_encoding_doc_v2_distill = ModelMeta(
     max_tokens=512,
     reference="https://huggingface.co/opensearch-project/opensearch-neural-sparse-encoding-doc-v2-distill",
     similarity_fn_name="dot",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     public_training_code="https://github.com/zhichao-aws/opensearch-sparse-model-tuning-sample",
     public_training_data=True,
     use_instructions=True,
@@ -211,7 +211,7 @@ opensearch_neural_sparse_encoding_doc_v2_mini = ModelMeta(
     max_tokens=512,
     reference="https://huggingface.co/opensearch-project/opensearch-neural-sparse-encoding-doc-v2-mini",
     similarity_fn_name="dot",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     public_training_code="https://github.com/zhichao-aws/opensearch-sparse-model-tuning-sample",
     public_training_data=True,
     use_instructions=True,
@@ -233,7 +233,7 @@ opensearch_neural_sparse_encoding_doc_v1 = ModelMeta(
     max_tokens=512,
     reference="https://huggingface.co/opensearch-project/opensearch-neural-sparse-encoding-doc-v1",
     similarity_fn_name="dot",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     public_training_code="https://github.com/zhichao-aws/opensearch-sparse-model-tuning-sample",
     public_training_data=True,
     use_instructions=True,

mteb/models/model_implementations/ops_moa_models.py

@@ -67,7 +67,7 @@ ops_moa_yuan_embedding = ModelMeta(
     open_weights=True,
     public_training_code=None,
     public_training_data=None,
-    framework=["PyTorch", "Sentence Transformers"],
+    framework=["PyTorch", "Sentence Transformers", "safetensors"],
     reference="https://huggingface.co/OpenSearch-AI/Ops-MoA-Yuan-embedding-1.0",
     similarity_fn_name="cosine",
     use_instructions=False,

mteb/models/model_implementations/ordalietech_solon_embeddings_mini_beta_1_1.py

@@ -14,7 +14,7 @@ solon_embeddings_1_1 = ModelMeta(
     max_tokens=8192,
     reference="https://huggingface.co/OrdalieTech/Solon-embeddings-mini-beta-1.1",
     similarity_fn_name="cosine",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors"],
     use_instructions=False,
     public_training_data=(
         "https://huggingface.co/datasets/PleIAs/common_corpus; "

mteb/models/model_implementations/pawan_models.py

@@ -26,7 +26,7 @@ pawan_embd_68m = ModelMeta(
     max_tokens=512,
     reference="https://huggingface.co/dmedhi/PawanEmbd-68M",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors"],
     adapted_from="ibm-granite/granite-embedding-278m-multilingual",
     superseded_by=None,
     public_training_code=None,

mteb/models/model_implementations/piccolo_models.py

@@ -18,7 +18,7 @@ piccolo_base_zh = ModelMeta(
     max_tokens=512,
     reference="https://huggingface.co/sensenova/piccolo-base-zh",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "Transformers"],
     use_instructions=False,
     superseded_by=None,
     adapted_from=None,

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

@@ -90,7 +90,7 @@ promptriever_llama2 = ModelMeta(
     ),
     reference="https://huggingface.co/samaya-ai/promptriever-llama2-7b-v1",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["PyTorch", "Tevatron"],
+    framework=["PyTorch", "Tevatron", "safetensors"],
     use_instructions=True,
     citation=PROMPTRIEVER_CITATION,
     public_training_code=None,
@@ -123,7 +123,7 @@ promptriever_llama3 = ModelMeta(
     license="apache-2.0",
     reference="https://huggingface.co/samaya-ai/promptriever-llama3.1-8b-v1",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["PyTorch", "Tevatron"],
+    framework=["PyTorch", "Tevatron", "safetensors"],
     use_instructions=True,
     citation=PROMPTRIEVER_CITATION,
     public_training_code=None,
@@ -156,7 +156,7 @@ promptriever_llama3_instruct = ModelMeta(
     license="apache-2.0",
     reference="https://huggingface.co/samaya-ai/promptriever-llama3.1-8b-instruct-v1",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["PyTorch", "Tevatron"],
+    framework=["PyTorch", "Tevatron", "safetensors"],
     use_instructions=True,
     citation=PROMPTRIEVER_CITATION,
     public_training_code=None,
@@ -189,7 +189,7 @@ promptriever_mistral_v1 = ModelMeta(
     license="apache-2.0",
     reference="https://huggingface.co/samaya-ai/promptriever-mistral-v0.1-7b-v1",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["PyTorch", "Tevatron"],
+    framework=["PyTorch", "Tevatron", "safetensors"],
     use_instructions=True,
     citation=PROMPTRIEVER_CITATION,
     public_training_code=None,

mteb/models/model_implementations/pylate_models.py

@@ -19,6 +19,7 @@ from mteb.types import (
     Array,
     BatchedInput,
     CorpusDatasetType,
+    EncodeKwargs,
     PromptType,
     QueryDatasetType,
     RetrievalOutputType,
@@ -29,7 +30,7 @@ logger = logging.getLogger(__name__)
 
 
 class PylateSearchEncoder:
-    """Mixin class to add PyLate-based indexing and search to an encoder. Implements :class:`SearchProtocol`"""
+    """Mixin class to add PyLate-based indexing and search to an encoder. Implements [SearchProtocol][mteb.models.SearchProtocol]"""
 
     base_index_dir: Path | None = None
     _index_dir: Path | None = None
@@ -45,7 +46,7 @@ class PylateSearchEncoder:
         task_metadata: TaskMetadata,
         hf_split: str,
         hf_subset: str,
-        encode_kwargs: dict[str, Any],
+        encode_kwargs: EncodeKwargs,
     ) -> None:
         """Index the corpus for retrieval.
 
@@ -78,7 +79,7 @@ class PylateSearchEncoder:
         hf_split: str,
         hf_subset: str,
         top_k: int,
-        encode_kwargs: dict[str, Any],
+        encode_kwargs: EncodeKwargs,
         top_ranked: TopRankedDocumentsType | None = None,
     ) -> RetrievalOutputType:
         queries_dataloader = create_dataloader(
@@ -136,7 +137,7 @@ class PylateSearchEncoder:
         hf_subset: str,
         hf_split: str,
         top_k: int,
-        encode_kwargs: dict[str, Any],
+        encode_kwargs: EncodeKwargs,
     ) -> dict[str, list[tuple[float, str]]]:
         from pylate import indexes, retrieve
 
@@ -200,7 +201,7 @@ class PylateSearchEncoder:
         task_metadata: TaskMetadata,
         hf_subset: str,
         hf_split: str,
-        encode_kwargs: dict[str, Any],
+        encode_kwargs: EncodeKwargs,
     ) -> dict[str, list[tuple[float, str]]]:
         """Rerank with PyLate's rank.rerank using per-query candidates.
 
@@ -350,7 +351,7 @@ colbert_v2 = ModelMeta(
     embed_dim=None,
     license="mit",
     similarity_fn_name=ScoringFunction.MAX_SIM,
-    framework=["PyLate", "ColBERT"],
+    framework=["PyLate", "ColBERT", "Transformers", "ONNX", "safetensors"],
     reference="https://huggingface.co/colbert-ir/colbertv2.0",
     use_instructions=False,
     adapted_from=None,
@@ -406,7 +407,7 @@ jina_colbert_v2 = ModelMeta(
     embed_dim=None,
     license="cc-by-nc-4.0",
     similarity_fn_name=ScoringFunction.MAX_SIM,
-    framework=["PyLate", "ColBERT"],
+    framework=["PyLate", "ColBERT", "ONNX", "safetensors"],
     reference="https://huggingface.co/jinaai/jina-colbert-v2",
     use_instructions=False,
     adapted_from=None,
@@ -439,7 +440,7 @@ jina_colbert_v2 = ModelMeta(
     url = "https://aclanthology.org/2024.mrl-1.11/",
     doi = "10.18653/v1/2024.mrl-1.11",
     pages = "159--166",
-    abstract = "Multi-vector dense models, such as ColBERT, have proven highly effective in information retrieval. ColBERT`s late interaction scoring approximates the joint query-document attention seen in cross-encoders while maintaining inference efficiency closer to traditional dense retrieval models, thanks to its bi-encoder architecture and recent optimizations in indexing and search. In this paper, we introduce a novel architecture and a training framework to support long context window and multilingual retrieval. Leveraging Matryoshka Representation Loss, we further demonstrate that the reducing the embedding dimensionality from 128 to 64 has insignificant impact on the model`s retrieval performance and cut storage requirements by up to 50{\%}. Our new model, Jina-ColBERT-v2, demonstrates strong performance across a range of English and multilingual retrieval tasks,"
+    abstract = "Multi-vector dense models, such as ColBERT, have proven highly effective in information retrieval. ColBERT`s late interaction scoring approximates the joint query-document attention seen in cross-encoders while maintaining inference efficiency closer to traditional dense retrieval models, thanks to its bi-encoder architecture and recent optimizations in indexing and search. In this paper, we introduce a novel architecture and a training framework to support long context window and multilingual retrieval. Leveraging Matryoshka Representation Loss, we further demonstrate that the reducing the embedding dimensionality from 128 to 64 has insignificant impact on the model`s retrieval performance and cut storage requirements by up to 50{\\%}. Our new model, Jina-ColBERT-v2, demonstrates strong performance across a range of English and multilingual retrieval tasks,"
 }""",
 )
 
@@ -462,7 +463,7 @@ lightonai__gte_moderncolbert_v1 = ModelMeta(
     embed_dim=None,
     license="apache-2.0",
     similarity_fn_name="MaxSim",
-    framework=["PyLate", "ColBERT"],
+    framework=["PyLate", "ColBERT", "safetensors", "Sentence Transformers"],
     reference="https://huggingface.co/lightonai/GTE-ModernColBERT-v1",
     use_instructions=False,
     adapted_from="Alibaba-NLP/gte-modernbert-base",

mteb/models/model_implementations/qodo_models.py

@@ -42,7 +42,7 @@ Qodo_Embed_1_1_5B = ModelMeta(
     max_tokens=32768,
     reference="https://huggingface.co/Qodo/Qodo-Embed-1-1.5B",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     use_instructions=False,
     public_training_code=None,
     public_training_data=None,
@@ -65,7 +65,7 @@ Qodo_Embed_1_7B = ModelMeta(
     max_tokens=32768,
     reference="https://huggingface.co/Qodo/Qodo-Embed-1-7B",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     use_instructions=False,
     public_training_code=None,
     public_training_data=None,

mteb/models/model_implementations/qtack_models.py

@@ -37,7 +37,7 @@ mini_gte = ModelMeta(
     max_tokens=512,
     reference="https://huggingface.co/prdev/mini-gte",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors"],
     use_instructions=False,
     public_training_code=None,
     public_training_data=None,

mteb/models/model_implementations/qwen3_models.py

@@ -146,7 +146,7 @@ Qwen3_Embedding_0B6 = ModelMeta(
     license="apache-2.0",
     reference="https://huggingface.co/Qwen/Qwen3-Embedding-0.6B",
     similarity_fn_name="cosine",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     use_instructions=True,
     public_training_code=None,
     public_training_data=None,
@@ -169,7 +169,7 @@ Qwen3_Embedding_4B = ModelMeta(
     license="apache-2.0",
     reference="https://huggingface.co/Qwen/Qwen3-Embedding-4B",
     similarity_fn_name="cosine",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     use_instructions=True,
     public_training_code=None,
     public_training_data=None,
@@ -192,7 +192,7 @@ Qwen3_Embedding_8B = ModelMeta(
     license="apache-2.0",
     reference="https://huggingface.co/Qwen/Qwen3-Embedding-8B",
     similarity_fn_name="cosine",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     use_instructions=True,
     public_training_code=None,
     public_training_data=None,

mteb/models/model_implementations/qzhou_models.py

@@ -70,7 +70,7 @@ QZhou_Embedding = ModelMeta(
     max_tokens=8192,
     reference="https://huggingface.co/Kingsoft-LLM/QZhou-Embedding",
     similarity_fn_name="cosine",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors"],
     use_instructions=True,
     public_training_code=None,
     public_training_data="https://huggingface.co/datasets/cfli/bge-full-data",
@@ -104,7 +104,7 @@ QZhou_Embedding_Zh = ModelMeta(
     max_tokens=8192,
     reference="http://huggingface.co/Kingsoft-LLM/QZhou-Embedding-Zh",
     similarity_fn_name="cosine",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors"],
     use_instructions=True,
     public_training_code=None,
     public_training_data=None,

mteb/models/model_implementations/random_baseline.py

@@ -68,7 +68,7 @@ _common_mock_metadata = dict(
     license="mit",
     max_tokens=np.inf,
     reference=None,
-    similarity_fn_name="cosine",  # type: ignore
+    similarity_fn_name="cosine",
     framework=[],
     use_instructions=False,
     public_training_code=None,  # No training code, as this is a random baseline
@@ -187,7 +187,7 @@ class RandomEncoderBaseline:
 
 
 random_encoder_baseline = ModelMeta(
-    loader=RandomEncoderBaseline,  # type: ignore
+    loader=RandomEncoderBaseline,
     name="baseline/random-encoder-baseline",
     model_type=["dense"],
     modalities=["text", "image"],
@@ -232,7 +232,7 @@ class RandomCrossEncoderBaseline:
 
 
 random_cross_encoder_baseline = ModelMeta(
-    loader=RandomCrossEncoderBaseline,  # type: ignore
+    loader=RandomCrossEncoderBaseline,
     name="baseline/random-cross-encoder-baseline",
     model_type=["cross-encoder"],
     modalities=["text", "image"],

mteb/models/model_implementations/rasgaard_models.py

@@ -4,7 +4,7 @@ from mteb.models.model_implementations.model2vec_models import Model2VecModel
 from mteb.models.model_meta import ModelMeta, ScoringFunction
 
 potion_base_8m = ModelMeta(
-    loader=Model2VecModel,  # type: ignore
+    loader=Model2VecModel,
     name="rasgaard/m2v-dfm-large",
     model_type=["dense"],
     languages=["dan-Latn"],
@@ -17,7 +17,7 @@ potion_base_8m = ModelMeta(
     embed_dim=256,
     license="mit",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["NumPy", "Sentence Transformers"],
+    framework=["NumPy", "Sentence Transformers", "safetensors"],
     reference="https://huggingface.co/rasgaard/m2v-dfm-large",
     use_instructions=False,
     adapted_from="KennethEnevoldsen/dfm-sentence-encoder-large",

mteb/models/model_implementations/reasonir_model.py

@@ -56,7 +56,7 @@ ReasonIR_8B = ModelMeta(
     max_tokens=131072,
     reference="https://huggingface.co/ReasonIR/ReasonIR-8B",
     similarity_fn_name="cosine",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "Transformers", "safetensors"],
     use_instructions=True,
     training_datasets=REASONIR_TRAINING_DATA,
     public_training_code="https://github.com/facebookresearch/ReasonIR/tree/main/training",

mteb/models/model_implementations/repllama_models.py

@@ -154,7 +154,7 @@ REPLLAMA_CITATION = """
 """
 
 repllama_llama2_original = ModelMeta(
-    loader=RepLLaMAModel,  # type: ignore
+    loader=RepLLaMAModel,
     loader_kwargs=dict(
         base_model_name_or_path="meta-llama/Llama-2-7b-hf",
         device_map="auto",
@@ -187,7 +187,7 @@ repllama_llama2_original = ModelMeta(
 
 
 repllama_llama2_reproduced = ModelMeta(
-    loader=RepLLaMAModel,  # type: ignore
+    loader=RepLLaMAModel,
     loader_kwargs=dict(
         base_model_name_or_path="meta-llama/Llama-2-7b-hf",
         device_map="auto",
@@ -207,7 +207,7 @@ repllama_llama2_reproduced = ModelMeta(
     license="apache-2.0",
     reference="https://huggingface.co/samaya-ai/RepLLaMA-reproduced",
     similarity_fn_name=ScoringFunction.COSINE,
-    framework=["PyTorch", "Tevatron"],
+    framework=["PyTorch", "Tevatron", "safetensors"],
     use_instructions=True,
     citation=REPLLAMA_CITATION,
     public_training_code=None,

mteb/models/model_implementations/rerankers_custom.py

@@ -214,7 +214,7 @@ class JinaReranker(RerankerWrapper):
 
 
 monobert_large = ModelMeta(
-    loader=MonoBERTReranker,  # type: ignore
+    loader=MonoBERTReranker,
     loader_kwargs=dict(
         fp_options="float16",
     ),
@@ -234,12 +234,12 @@ monobert_large = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "Transformers"],
 )
 
 # languages unclear: https://huggingface.co/jinaai/jina-reranker-v2-base-multilingual/discussions/28
 jina_reranker_multilingual = ModelMeta(
-    loader=JinaReranker,  # type: ignore
+    loader=JinaReranker,
     loader_kwargs=dict(
         fp_options="float16",
     ),
@@ -259,11 +259,17 @@ jina_reranker_multilingual = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=[
+        "Sentence Transformers",
+        "PyTorch",
+        "Transformers",
+        "ONNX",
+        "safetensors",
+    ],
 )
 
 bge_reranker_v2_m3 = ModelMeta(
-    loader=BGEReranker,  # type: ignore
+    loader=BGEReranker,
     loader_kwargs=dict(
         fp_options="float16",
     ),
@@ -316,7 +322,7 @@ bge_reranker_v2_m3 = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=bge_m3_training_data,
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors", "Transformers"],
     citation="""
     @misc{li2023making,
       title={Making Large Language Models A Better Foundation For Dense Retrieval},

mteb/models/model_implementations/rerankers_monot5_based.py

@@ -330,7 +330,7 @@ monot5_small = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers"],
     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},
@@ -343,7 +343,7 @@ monot5_small = ModelMeta(
 )
 
 monot5_base = ModelMeta(
-    loader=MonoT5Reranker,  # type: ignore
+    loader=MonoT5Reranker,
     loader_kwargs=dict(
         fp_options="float16",
     ),
@@ -372,7 +372,7 @@ monot5_base = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers"],
 )
 
 monot5_large = ModelMeta(
@@ -396,7 +396,7 @@ monot5_large = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers"],
     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},
@@ -429,7 +429,7 @@ monot5_3b = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers"],
     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},
@@ -442,7 +442,7 @@ monot5_3b = ModelMeta(
 )
 
 flant5_base = ModelMeta(
-    loader=FLANT5Reranker,  # type: ignore
+    loader=FLANT5Reranker,
     loader_kwargs=dict(
         fp_options="float16",
     ),
@@ -484,7 +484,7 @@ flant5_base = ModelMeta(
     public_training_data=None,
     similarity_fn_name=None,
     use_instructions=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers", "safetensors"],
 )
 
 flant5_large = ModelMeta(
@@ -530,7 +530,7 @@ flant5_large = ModelMeta(
     public_training_data=None,
     similarity_fn_name=None,
     use_instructions=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers", "safetensors"],
 )
 
 flant5_xl = ModelMeta(
@@ -576,7 +576,7 @@ flant5_xl = ModelMeta(
     public_training_data=None,
     similarity_fn_name=None,
     use_instructions=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers", "safetensors"],
 )
 
 flant5_xxl = ModelMeta(
@@ -622,7 +622,7 @@ flant5_xxl = ModelMeta(
     public_training_data=None,
     similarity_fn_name=None,
     use_instructions=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers", "safetensors"],
 )
 
 
@@ -647,7 +647,7 @@ llama2_7b = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers", "safetensors"],
     citation="""@misc{touvron2023llama2openfoundation,
       title={Llama 2: Open Foundation and Fine-Tuned Chat Models},
       author={Hugo Touvron and Louis Martin and Kevin Stone and Peter Albert and Amjad Almahairi and Yasmine Babaei and Nikolay Bashlykov and Soumya Batra and Prajjwal Bhargava and Shruti Bhosale and Dan Bikel and Lukas Blecher and Cristian Canton Ferrer and Moya Chen and Guillem Cucurull and David Esiobu and Jude Fernandes and Jeremy Fu and Wenyin Fu and Brian Fuller and Cynthia Gao and Vedanuj Goswami and Naman Goyal and Anthony Hartshorn and Saghar Hosseini and Rui Hou and Hakan Inan and Marcin Kardas and Viktor Kerkez and Madian Khabsa and Isabel Kloumann and Artem Korenev and Punit Singh Koura and Marie-Anne Lachaux and Thibaut Lavril and Jenya Lee and Diana Liskovich and Yinghai Lu and Yuning Mao and Xavier Martinet and Todor Mihaylov and Pushkar Mishra and Igor Molybog and Yixin Nie and Andrew Poulton and Jeremy Reizenstein and Rashi Rungta and Kalyan Saladi and Alan Schelten and Ruan Silva and Eric Michael Smith and Ranjan Subramanian and Xiaoqing Ellen Tan and Binh Tang and Ross Taylor and Adina Williams and Jian Xiang Kuan and Puxin Xu and Zheng Yan and Iliyan Zarov and Yuchen Zhang and Angela Fan and Melanie Kambadur and Sharan Narang and Aurelien Rodriguez and Robert Stojnic and Sergey Edunov and Thomas Scialom},
@@ -689,7 +689,7 @@ llama2_7b_chat = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers", "safetensors"],
 )
 
 mistral_7b = ModelMeta(
@@ -713,7 +713,7 @@ mistral_7b = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers", "safetensors"],
     citation="""@misc{jiang2023mistral7b,
       title={Mistral 7B},
       author={Albert Q. Jiang and Alexandre Sablayrolles and Arthur Mensch and Chris Bamford and Devendra Singh Chaplot and Diego de las Casas and Florian Bressand and Gianna Lengyel and Guillaume Lample and Lucile Saulnier and Lélio Renard Lavaud and Marie-Anne Lachaux and Pierre Stock and Teven Le Scao and Thibaut Lavril and Thomas Wang and Timothée Lacroix and William El Sayed},
@@ -748,7 +748,7 @@ followir_7b = ModelMeta(
     public_training_data=None,
     similarity_fn_name=None,
     use_instructions=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers", "safetensors"],
     citation="""
     @misc{weller2024followir,
       title={FollowIR: Evaluating and Teaching Information Retrieval Models to Follow Instructions},
@@ -898,11 +898,11 @@ mt5_base_mmarco_v2 = ModelMeta(
     public_training_data=None,
     similarity_fn_name=None,
     use_instructions=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers"],
 )
 
 mt5_13b_mmarco_100k = ModelMeta(
-    loader=MonoT5Reranker,  # type: ignore
+    loader=MonoT5Reranker,
     loader_kwargs=dict(
         fp_options="float16",
     ),
@@ -922,5 +922,5 @@ mt5_13b_mmarco_100k = ModelMeta(
     similarity_fn_name=None,
     use_instructions=None,
     training_datasets=None,
-    framework=["PyTorch"],
+    framework=["PyTorch", "Transformers"],
 )

mteb/models/model_implementations/richinfoai_models.py

@@ -21,7 +21,7 @@ ritrieve_zh_v1 = ModelMeta(
     max_tokens=512,
     reference="https://huggingface.co/richinfoai/ritrieve_zh_v1",
     similarity_fn_name="cosine",
-    framework=["Sentence Transformers", "PyTorch"],
+    framework=["Sentence Transformers", "PyTorch", "safetensors"],
     use_instructions=False,
     superseded_by=None,
     adapted_from=None,