all-MiniLM-L6-v35-pair_score

This is a sentence-transformers model finetuned from KhaledReda/all-MiniLM-L6-v34-pair_score on the pairs_with_scores_v30 dataset. It maps sentences & paragraphs to a 384-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: KhaledReda/all-MiniLM-L6-v34-pair_score
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 384 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • pairs_with_scores_v30
• Language: en
• License: apache-2.0

Model Sources

• Documentation: Sentence Transformers Documentation
• Repository: Sentence Transformers on GitHub
• Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 256, 'do_lower_case': False, 'architecture': 'BertModel'})
  (1): Pooling({'word_embedding_dimension': 384, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
  (2): Normalize()
)

Usage

Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("sentence_transformers_model_id")
# Run inference
sentences = [
    'sanosan',
    'v-neck cotton vest black category fashion designer wear outerwear vest tags classic suit vest buttonup front pockets vest false welt pockets vest adjustable strap vest cotton vest vest vneck vest gilet vneck gilet keywords vest vneck vest gilet vneck gilet attrs gender women brand palma generic name vest size s features adjustable strap with buckle types of fashion styles classic formal collar v-neck closure style button-up pocket style false welt pockets sleeve style sleeveless material cotton color black fashion style v-neck description our v-neck cotton vest inspired by classic suit vests offers versatility and style. the button-up front and false welt pockets on the front add a touch of elegance while the adjustable strap with a buckle on the back ensures comfort and customization. whether for formal occasions or adding a polished touch to your everyday attire this refined cotton vest is a stylish choice and a perfect addition to your wardrobe. material 100 cotton. model is wearing size m.',
    'lash brow category beauty skincare eye treatment eye treatment tags lashes growth serum brow serum lash serum keywords brow serum lash serum description this natural growth serum causes existing lashes to become longer and stimulates growth in hair follicles not currently producing lashes. it has shown effects after 4 weeks of usage which is why this product is considered to be highly effective.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[ 1.0000, -0.0478, -0.1530],
#         [-0.0478,  1.0000, -0.0644],
#         [-0.1530, -0.0644,  1.0000]])

Training Details

Training Dataset

pairs_with_scores_v30

• Dataset: pairs_with_scores_v30 at 33eac44
• Size: 19,759,758 training samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 3 tokensmean: 5.54 tokensmax: 20 tokens	min: 13 tokensmean: 86.78 tokensmax: 256 tokens	min: 0.0mean: 0.04max: 1.0

• Samples:

  sentence1	sentence2	score
  boots 3780 nort	mediterranean blue mug category home and garden drinkware mug mug tags blue mug porcelain mug mediterranean mug mug keywords mediterranean mug mug attrs brand inches home generic name mug product name mediterranean measurements 350 ml number of pieces 1 shape mug purpose drinking material porcelain color blue description embrace the warmth and affection with mediterranean blue mug that makes your heartwarming moments even more special. product specifications material high quality porcelain.	0.0
  men shorts	ardell natural lashes black /125 category beauty cosmetics eye make-up eyelash tags ardell ardell lashes ardell natural lashes natural lashes keywords ardell ardell lashes ardell natural lashes natural lashes attrs color black	0.0
  boots 2316 b-16-j	fresh iced pink grapefruit juice category restaurants juices and drinks beverage juice tags fresh iced juice fresh juice iced juice juice pink grapefruit juice keywords fresh juice iced juice juice pink grapefruit juice	0.0

• Loss: CoSENTLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "pairwise_cos_sim"
  }
  

Evaluation Dataset

pairs_with_scores_v30

• Dataset: pairs_with_scores_v30 at 33eac44
• Size: 99,296 evaluation samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 3 tokensmean: 5.33 tokensmax: 19 tokens	min: 16 tokensmean: 86.78 tokensmax: 256 tokens	min: 0.0mean: 0.05max: 1.0

• Samples:

  sentence1	sentence2	score
  accuchek guide me	bright peach cover up category fashion swimwear bathing cover bathing cover tags peach cover up cover-up bright cover up cover up keywords bright cover up cover up attrs gender women brand pepla generic name cover up color bright peach	0.0
  medical accessory	ferragamo for men edt 100 ml category beauty fragrances perfume perfume tags men eau de toilette fragrances eau de toilette ferragamo ferragamo eau de toilette perfume cologne fragrance keywords eau de toilette ferragamo ferragamo eau de toilette perfume cologne fragrance attrs units 100 millilitre target group men	0.0
  muslin kimono	fake date cookies category restaurants keto sweet biscuit tags coconut flour cookies grass fed butter cookies ghee cookies cinnamon powder cookies raw cacao cookies sesame cookies low carb cookies diabetic friendly cookies keto friendly cookies gluten free cookies almond flour cookies cookies date cookies fake date cookies keywords almond flour cookies cookies date cookies fake date cookies description almond flour - coconut flour - grass fed butter - ghee - egg - cinnamon powder - raw cacao - sesame nutritional facts per 10 gr 58 kcal - 4.6 gr fat - 1 gr net carb - 1.2 gr fiber - 1.9 gr protein suitable for low carb - diabetic friendly - keto friendly - gluten free 14 day on shelf air tight container .	0.0

• Loss: CoSENTLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "pairwise_cos_sim"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 128
• per_device_eval_batch_size: 128
• learning_rate: 2e-05
• num_train_epochs: 1
• warmup_ratio: 0.1
• fp16: True

All Hyperparameters

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 128
• per_device_eval_batch_size: 128
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 2e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 1
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.1
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• save_safetensors: True
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• no_cuda: False
• use_cpu: False
• use_mps_device: False
• seed: 42
• data_seed: None
• jit_mode_eval: False
• use_ipex: False
• bf16: False
• fp16: True
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• local_rank: 0
• ddp_backend: None
• tpu_num_cores: None
• tpu_metrics_debug: False
• debug: []
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_prefetch_factor: None
• past_index: -1
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: False
• ignore_data_skip: False
• fsdp: []
• fsdp_min_num_params: 0
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• fsdp_transformer_layer_cls_to_wrap: None
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch
• optim_args: None
• adafactor: False
• group_by_length: False
• length_column_name: length
• ddp_find_unused_parameters: None
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• skip_memory_metrics: True
• use_legacy_prediction_loop: False
• push_to_hub: False
• resume_from_checkpoint: None
• hub_model_id: None
• hub_strategy: every_save
• hub_private_repo: None
• hub_always_push: False
• hub_revision: None
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• eval_do_concat_batches: True
• fp16_backend: auto
• push_to_hub_model_id: None
• push_to_hub_organization: None
• mp_parameters:
• auto_find_batch_size: False
• full_determinism: False
• torchdynamo: None
• ray_scope: last
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• include_tokens_per_second: False
• include_num_input_tokens_seen: False
• neftune_noise_alpha: None
• optim_target_modules: None
• batch_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• liger_kernel_config: None
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Framework Versions

• Python: 3.12.3
• Sentence Transformers: 5.1.0
• Transformers: 4.55.4
• PyTorch: 2.6.0+cu124
• Accelerate: 1.10.1
• Datasets: 4.0.0
• Tokenizers: 0.21.4

Citation

BibTeX

Sentence Transformers

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}

CoSENTLoss

@online{kexuefm-8847,
    title={CoSENT: A more efficient sentence vector scheme than Sentence-BERT},
    author={Su Jianlin},
    year={2022},
    month={Jan},
    url={https://kexue.fm/archives/8847},
}