Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/huggingface/sentence-transformers
/ functions
Functions
4,575 in github.com/huggingface/sentence-transformers
⨍
Functions
4,575
◇
Types & classes
516
↳
Endpoints
81
↓ 1 callers
Method
__init__
( self, dataset_folder, s1_col_idx=0, s2_col_idx=1, score_col_idx=2,
sentence_transformers/sentence_transformer/readers/sts_data.py:26
↓ 1 callers
Method
__init__
(self, fn)
sentence_transformers/multi_vector_encoder/losses/matryoshka.py:21
↓ 1 callers
Method
__init__
(self, fn)
sentence_transformers/multi_vector_encoder/losses/token_pooling.py:28
↓ 1 callers
Method
__init__
PListMLE loss for learning to rank with position-aware weighting. This loss function implements the ListMLE ranking algorithm which u
sentence_transformers/cross_encoder/losses/plist_mle.py:50
↓ 1 callers
Method
__init__
(self, linear: nn.Linear)
sentence_transformers/sparse_encoder/modules/sparse_auto_encoder.py:16
↓ 1 callers
Method
__init__
CSRLoss implements a combined loss function for Contrastive Sparse Representation (CSR) models. This loss combines two components:
sentence_transformers/sparse_encoder/losses/csr.py:140
↓ 1 callers
Method
__init__
Initialize a BaseModel instance. Args: model_name_or_path (str, optional): If a filepath on disk, loads the model from t
sentence_transformers/base/model.py:104
↓ 1 callers
Method
__init__
(self, unpicklable: bool = False)
tests/utils.py:52
↓ 1 callers
Method
__init__
(self, *args, **kwargs)
tests/sentence_transformer/test_trainer.py:488
↓ 1 callers
Method
__init__
(self)
tests/multi_vector_encoder/losses/test_token_pooling.py:301
↓ 1 callers
Method
__init__
(self, result: dict[str, float])
tests/base/test_model.py:604
↓ 1 callers
Method
__init__
(self, routed=False, buffers=False)
tests/base/test_gradcache_distributed.py:20
↓ 1 callers
Method
__init__
(self, tensor)
tests/base/modules/test_transformer.py:829
↓ 1 callers
Method
__len__
(self)
sentence_transformers/sentence_transformer/datasets/no_duplicates_dataloader.py:59
↓ 1 callers
Function
_add_output_column
(resolved: dict[str, Any], name: str, values: list)
sentence_transformers/util/dataset.py:148
↓ 1 callers
Method
_apply
Apply a function to every parameter and buffer, as :meth:`torch.nn.Module._apply` does. accelerate patches ``to`` and ``cuda`` on the backbon
sentence_transformers/base/model.py:1798
↓ 1 callers
Method
_apply_legacy_fixups
Patch up modules loaded from save formats that predate :class:`MultiVectorEncoder` (PyLate v3, Stanford-NLP ColBERT). Each step is a no-op for
sentence_transformers/multi_vector_encoder/model.py:1168
↓ 1 callers
Function
_assert_output_equal
(actual, expected)
tests/base/test_distributed_evaluation.py:48
↓ 1 callers
Function
_batch_singular_input
( embeddings: list | np.ndarray | Tensor, mask: Tensor | None )
sentence_transformers/util/similarity.py:277
↓ 1 callers
Function
_build_mako_lut
(n: int = 256)
sentence_transformers/multi_vector_encoder/interpretability.py:37
↓ 1 callers
Method
_build_stanford_projection
Build a token-level :class:`~sentence_transformers.base.modules.dense.Dense` from a Stanford-NLP ColBERT checkpoint. Stanford-NLP che
sentence_transformers/multi_vector_encoder/model.py:1443
↓ 1 callers
Function
_cached_mnrl
(model: nn.Module, **kwargs)
tests/multi_vector_encoder/losses/test_token_pooling.py:57
↓ 1 callers
Method
_calculate_avg_flops
Calculate average FLOPS between queries and corpus. FLOPS is the dot product of the average query and corpus count vectors, and can b
sentence_transformers/sparse_encoder/evaluation/sparse_information_retrieval.py:304
↓ 1 callers
Method
_calculate_weight_value
Calculate the weight value based on the current step and scheduler type.
sentence_transformers/sparse_encoder/callbacks/splade_callbacks.py:99
↓ 1 callers
Method
_call_multimodal_processor
Call a :class:`ProcessorMixin` processor, handling both legacy and v5 calling conventions.
sentence_transformers/base/modules/transformer.py:1831
↓ 1 callers
Method
_call_single_modality_processor
Call a single-modality processor (tokenizer, feature extractor, image/video processor).
sentence_transformers/base/modules/transformer.py:1877
↓ 1 callers
Method
_chat_template_declares_task
Whether the active chat template (the processor's, else the tokenizer's) uses a ``task`` variable, e.g. to append query augmentation tokens fo
sentence_transformers/base/modules/transformer.py:1227
↓ 1 callers
Method
_checkpoint_push_skip_names
(self, checkpoint_folder: str)
sentence_transformers/base/trainer.py:1060
↓ 1 callers
Function
_chunked_outputs
Run ``model`` over row chunks of ``features`` and return one output dict per chunk. A single-element result means the batch ran in one forward, i
sentence_transformers/base/losses/merged_forward.py:126
↓ 1 callers
Function
_clear_warning_once_cache
()
tests/conftest.py:29
↓ 1 callers
Function
_colbert_kd_scores
Score each query's candidate group with the supplied pairwise callable.
sentence_transformers/multi_vector_encoder/scoring/colbert.py:69
↓ 1 callers
Function
_colbert_scores
Score every query against all candidate groups with the supplied callable.
sentence_transformers/multi_vector_encoder/scoring/colbert.py:197
↓ 1 callers
Function
_colbert_scores_pairwise
Translate training mask names for the supplied pairwise callable.
sentence_transformers/multi_vector_encoder/scoring/colbert.py:138
↓ 1 callers
Method
_compute_column_stats
Compute the stats entry for a single column using bounded-memory chunking. Iterates at most 100 rows in chunks of 8 so peak memory is
sentence_transformers/base/model_card.py:1013
↓ 1 callers
Function
_config_diff
(reference: dict[str, Any], config: dict[str, Any], index: int)
sentence_transformers/base/merging.py:24
↓ 1 callers
Method
_configure_clustering
(self, device: torch.device)
sentence_transformers/multi_vector_encoder/modules/token_pooling.py:397
↓ 1 callers
Function
_create_kernel_mapping
(config: KernelConfig, device: torch.device)
sentence_transformers/util/kernels.py:39
↓ 1 callers
Function
_ddp_replay_context
(model: Any, synchronize: bool)
sentence_transformers/base/losses/gradcache.py:295
↓ 1 callers
Method
_determine_labels_to_use
(dataset: Dataset, valid_label_columns: list[str] | None)
sentence_transformers/base/sampler.py:305
↓ 1 callers
Function
_embeddings
(count: int = 3)
tests/sparse_encoder/losses/test_sparse_distill_kl_div.py:10
↓ 1 callers
Method
_eval_during_training
Runs evaluation during the training
sentence_transformers/sentence_transformer/fit_mixin.py:702
↓ 1 callers
Method
_eval_during_training
Runs evaluation during the training
sentence_transformers/cross_encoder/fit_mixin.py:571
↓ 1 callers
Method
_example_asset_html
Generate an inline HTML tag for a saved example asset.
sentence_transformers/base/model_card.py:1830
↓ 1 callers
Method
_exclude_prompt_from_mask
Zero out prompt token positions in the attention mask so they are excluded from pooling. Operates on a copy: the gradient-cached losses re-em
sentence_transformers/sentence_transformer/modules/pooling.py:154
↓ 1 callers
Function
_features
(rank, case)
tests/base/test_gradcache_distributed.py:59
↓ 1 callers
Method
_flattened
(self, lengths: list[int])
tests/sentence_transformer/losses/test_gradcache.py:273
↓ 1 callers
Function
_format_bytes
Human-readable byte formatting.
examples/sentence_transformer/evaluation/evaluation_no_dup_batch_sampler_speed.py:214
↓ 1 callers
Method
_forward_flattened
( self, token_embeddings: Tensor, features: dict[str, Tensor], prompt_length:
sentence_transformers/sentence_transformer/modules/pooling.py:257
↓ 1 callers
Method
_forward_flattened
(self, features: dict[str, torch.Tensor])
sentence_transformers/sparse_encoder/modules/splade_pooling.py:143
↓ 1 callers
Method
_forward_padded
( self, token_embeddings: Tensor, attention_mask: Tensor, features: dict[str,
sentence_transformers/sentence_transformer/modules/pooling.py:171
↓ 1 callers
Method
_generate_text_snippet
Generate a text-only usage snippet.
sentence_transformers/base/model_card.py:1692
↓ 1 callers
Method
_get_human_readable_name
(self, dataset_name: DatasetNameType | str)
sentence_transformers/sentence_transformer/evaluation/nano_beir.py:435
↓ 1 callers
Method
_get_human_readable_name
(self, dataset_name: DatasetNameType | str)
sentence_transformers/cross_encoder/evaluation/nano_beir.py:327
↓ 1 callers
Method
_get_max_length_for_task
Resolve the training max-length override for a column: an int applies to every column, a dict is keyed by task.
sentence_transformers/base/data_collator.py:88
↓ 1 callers
Method
_get_model_config
(self)
sentence_transformers/cross_encoder/model.py:865
↓ 1 callers
Method
_get_module_init_defaults
Hook for subclasses to inject extra defaults into a module's ``__init__`` at load time. Returned kwargs are forwarded to ``module_class.load(
sentence_transformers/base/model.py:1598
↓ 1 callers
Method
_get_neuron_worker_cores
Validate Neuron pool allocation and map requested devices to visible cores.
sentence_transformers/base/model.py:1951
↓ 1 callers
Function
_get_source_ref
()
docs/conf.py:109
↓ 1 callers
Method
_get_splade_pooling
Returns the SpladePooling module if present, or None. Only searches top-level modules.
sentence_transformers/sparse_encoder/model.py:1285
↓ 1 callers
Method
_get_worker_result
Wait for a result, checking for dead workers whenever the queue is empty.
sentence_transformers/base/model.py:2079
↓ 1 callers
Function
_grouped_barplot_ratios
(group_name_to_x_to_y: dict[str, dict[int, float]], ax: plt.Axes | None = None)
examples/sentence_transformer/training/matryoshka/matryoshka_eval_stsb.py:20
↓ 1 callers
Function
_has_lowercase
Check whether a tokenizers normalizer (or sequence of normalizers) includes Lowercase.
sentence_transformers/base/modules/transformer.py:519
↓ 1 callers
Function
_hierarchical_linkage
(embedding: Tensor)
sentence_transformers/multi_vector_encoder/modules/token_pooling.py:271
↓ 1 callers
Function
_infer_id_col
(dataset: Dataset, input_col: str)
sentence_transformers/util/dataset.py:21
↓ 1 callers
Function
_infer_value_col
(dataset: Dataset, id_col: str, input_col: str)
sentence_transformers/util/dataset.py:32
↓ 1 callers
Method
_inference
Run local inference on normalized inputs with resolved arguments.
sentence_transformers/sentence_transformer/model.py:935
↓ 1 callers
Method
_inference
Run local inference on normalized inputs with resolved arguments.
sentence_transformers/multi_vector_encoder/model.py:946
↓ 1 callers
Method
_inference
Run local inference on normalized inputs with resolved arguments.
sentence_transformers/cross_encoder/model.py:671
↓ 1 callers
Method
_inference
Run local inference on normalized inputs with resolved arguments.
sentence_transformers/sparse_encoder/model.py:767
↓ 1 callers
Function
_is_string_encoded_list
(value: Any)
sentence_transformers/util/dataset.py:52
↓ 1 callers
Method
_iter_one_pass
Yield batches from a single shuffle of the dataset without resampling.
sentence_transformers/base/sampler.py:613
↓ 1 callers
Function
_iter_texts
Normalize a value into a list of strings for counting.
examples/sentence_transformer/evaluation/evaluation_no_dup_batch_sampler_speed.py:207
↓ 1 callers
Method
_load_best_model
(self)
sentence_transformers/base/trainer.py:702
↓ 1 callers
Method
_load_config
Loads the transformers configuration, resolving the base model's config for PEFT checkpoints Args: model_name_or_path (str): The
sentence_transformers/base/modules/transformer.py:2262
↓ 1 callers
Method
_load_converted_modules
Load a save of a different model type by building this class's default modules on top of it. The source's ``config_sentence_transformers.json
sentence_transformers/base/model.py:1518
↓ 1 callers
Method
_load_dataset
( self, dataset_name: DatasetNameType | str, **ir_evaluator_kwargs )
sentence_transformers/sentence_transformer/evaluation/nano_beir.py:442
↓ 1 callers
Method
_load_dataset
( self, dataset_name: DatasetNameType | str, **ir_evaluator_kwargs )
sentence_transformers/cross_encoder/evaluation/nano_beir.py:330
↓ 1 callers
Method
_load_dataset_subset_split
(self, subset: str, split: str, required_columns: list[str])
sentence_transformers/sentence_transformer/evaluation/nano_beir.py:481
↓ 1 callers
Method
_load_dataset_subset_split
(self, subset: str, split: str, required_columns: list[str])
sentence_transformers/cross_encoder/evaluation/nano_beir.py:389
↓ 1 callers
Method
_load_encoder_only_model
Load encoder-only variants for encoder-decoder architectures. Checks :data:`_ENCODER_ONLY_MODELS` for standard mappings and handles a few spe
sentence_transformers/base/modules/transformer.py:2407
↓ 1 callers
Function
_load_hf_dataset
Load a HF dataset split with an optional subset/config.
examples/sentence_transformer/evaluation/evaluation_no_dup_batch_sampler_speed.py:410
↓ 1 callers
Method
_load_model
Loads the transformers or PEFT model into the `auto_model` attribute Args: model_name_or_path (str): The model name on Hugging Fa
sentence_transformers/base/modules/transformer.py:2351
↓ 1 callers
Method
_load_modules
( self, model_name_or_path: str, token: bool | str | None, cache_folder: str |
sentence_transformers/base/model.py:1208
↓ 1 callers
Function
_loss_and_grads
( model: SentenceTransformer, use_mini_batched_version: bool, mini_batch_size: int, batch_size: int )
tests/sentence_transformer/losses/test_mega_batch_margin.py:21
↓ 1 callers
Function
_loss_and_grads
(model: CrossEncoder, loss_fn: torch.nn.Module, seed: int = 12)
tests/cross_encoder/losses/test_cached_multiple_negatives_ranking.py:18
↓ 1 callers
Function
_loss_and_grads
( model: SparseEncoder, loss_fn: torch.nn.Module, batch_size: int = 6 )
tests/sparse_encoder/losses/test_cached_splade.py:30
↓ 1 callers
Function
_make_random_image
(seed: int, size: int = 32)
tests/multi_vector_encoder/test_model.py:1793
↓ 1 callers
Function
_make_relative_loss
relative-margin GISTEmbedLoss with no gather/contrast, fed precomputed embeddings (#3819).
tests/sentence_transformer/losses/test_gist_embed.py:114
↓ 1 callers
Function
_make_relative_loss
relative-margin CachedGISTEmbedLoss fed precomputed embeddings (#3819).
tests/sentence_transformer/losses/test_cached_gist_embed.py:228
↓ 1 callers
Method
_masked_maximum
(data: Tensor, mask: Tensor, dim: int = 1)
sentence_transformers/sentence_transformer/losses/batch_semi_hard_triplet.py:161
↓ 1 callers
Method
_merge_processing_kwargs
Merge per-call ``processing_kwargs`` on top of the instance-level ``self.processing_kwargs``. The merge is shallow per top-level key: for eac
sentence_transformers/base/modules/transformer.py:1610
↓ 1 callers
Function
_negative_score_features
anchor0's positive has a negative cosine (-0.50). A non-paired candidate (column 1) is *more* similar (-0.49). anchor1 is paired with that candida
tests/sentence_transformer/losses/test_gist_embed.py:134
↓ 1 callers
Function
_negative_score_reps
anchor0's positive has a negative cosine (-0.50). A non-paired candidate (column 1) is *more* similar (-0.49). anchor1 is paired with that candida
tests/sentence_transformer/losses/test_cached_gist_embed.py:245
↓ 1 callers
Function
_normalize_query_expansion
Validate the dict, fill defaults, and return a fully-populated config (or ``None`` for the no-expansion case).
sentence_transformers/base/modules/transformer.py:607
↓ 1 callers
Function
_onnx_prepare_external_data
Point an ONNX model over the 2GB protobuf limit at its weights under the name Optimum fetches. Such a model keeps its weights in files besid
sentence_transformers/backend/utils.py:140
↓ 1 callers
Function
_output_col
(input_col: str, plural: bool = False)
sentence_transformers/util/dataset.py:43
↓ 1 callers
Method
_pad_features_for_hpu
Pad input features to the next power of 2 for HPU graph compatibility.
sentence_transformers/sentence_transformer/model.py:1017
↓ 1 callers
Function
_pad_multi_vector_inputs
Pad a list of variable-length multi-vector tensors into a single 3D tensor with a mask. Returns the padded tensor and either the user-provided ma
sentence_transformers/util/similarity.py:785
↓ 1 callers
Function
_pad_tensor_list
Stack per-item ``(tokens_i, dim)`` embeddings into a padded ``(batch, max_tokens, dim)``. ``pad_sequence`` copies item by item, which on an accel
sentence_transformers/util/similarity.py:760
↓ 1 callers
Function
_pairs
Embeddings whose dot-product similarity is exactly `scores`, one per input pair.
tests/sentence_transformer/losses/test_cosent.py:21
← previous
next →
601–700 of 4,575, ranked by callers