SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L6-v2. 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: sentence-transformers/all-MiniLM-L6-v2
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 384 dimensions
• Similarity Function: Cosine Similarity

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}) with Transformer model: 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("PeYing/model1_v2")
# Run inference
sentences = [
    'Can employees wear traditional attire?',
    '1. 1. Yes, acceptable traditional attire includes: \n2. 1. Malaysian Traditional Attire. \n3. 2.Malay Baju Kurung. \n4. 3. Baju Melayu for Muslim men. \n5. 4.Indian Saree. \n6. 5. Punjabi Suit. \n7. Chinese Cheongsam are acceptable.',
    '1. N03 : Monday to Friday, 8am to 5:30pm.\n2. N04 : Tuesday to Saturday, 8am to 5:30pm.\n3. N05 : Monday to Friday, 8:30am to 6pm.\n4. N06 : Monday to Friday, 9am to 6:30pm.\n5. N07 : Tuesday to Saturday, 8:30am to 6pm.\n6. N08 : Tuesday to Saturday, 9am to 6.30pm.\n7. N6 : Tuesday to Saturday, 8:30pm to 6:15pm.\n8. N9: 5 working days 2 days off, 7:30am to 5:15pm , 10:30am to 8:15pm.\n9. N10: 5 working days 2 days off, 10:30am to 8:15pm , 7:30am to 5:15pm.\n10. AA/BB/CC/A/B/C : 4 working days 2 days off, 6:30am to 6:30pm , 6:30pm to 6:30am.\n11. AA1/BB1/CC1/A1/B1/C1 : 4 working days 2 days off, 6:30am to 6:30pm , 6:30pm to 6:30am.\n12. GG/HH/II/GG1/HH1/II1 : 4 working days 2 days off, 7:30am to 7:30pm , 7:30pm to 7:30am.\n13. P1 : Monday to Thursday (4 working days 2 days off), 6:30am to 6:30pm , 6:30pm to 6:30am.\n14. P2 : Tuesday to Friday (4 working days 2 days off), 6:30am to 6:30pm , 6:30pm to 6:30am. \n15. U1/U2/U3/UU1/UU2/UU3 : 4 working days 2 days off, 7:30am to 7.30pm. \n16. V1/V2/V3/VV1/VV2/VV3 : 4 working days 2 days off, 8.30am to 8.30pm. \n17. W1/W2/W3/WW1/WW2/WW3 : 4 working days 2 days off, 6.30am to 6.30pm. \n18. H1 : Monday to Thursday (4 working days 2 days off), 6.30am to 6.30pm. \n19. H2 : Tuesday to Friday (4 working days 2 days off), 6.30am to 6.30pm. \n20. H3 : Wednesday to Saturday (4 working days 2 days off), 6.30am to 6.30pm. \n21. H6(applicable in S only) : Monday to Thursday (4 working days 2 days off), 7.30am to 7.30pm. \n22. H6(applicable in M only) : Monday to Thursday (4 working days 2 days off), 7.30am to 7.30pm.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Training Details

Training Dataset

Unnamed Dataset

• Size: 182 training samples
• Columns: sentence_0, sentence_1, and label
• Approximate statistics based on the first 182 samples:

  	sentence_0	sentence_1	label
  type	string	string	int
  details	min: 7 tokensmean: 14.43 tokensmax: 36 tokens	min: 5 tokensmean: 53.8 tokensmax: 256 tokens	1: 100.00%

• Samples:

  sentence_0	sentence_1	label
  List out all the work schedule for Carsem.	1. N03 : Monday to Friday, 8am to 5:30pm.2. N04 : Tuesday to Saturday, 8am to 5:30pm.3. N05 : Monday to Friday, 8:30am to 6pm.4. N06 : Monday to Friday, 9am to 6:30pm.5. N07 : Tuesday to Saturday, 8:30am to 6pm.6. N08 : Tuesday to Saturday, 9am to 6.30pm.7. N6 : Tuesday to Saturday, 8:30pm to 6:15pm.8. N9: 5 working days 2 days off, 7:30am to 5:15pm , 10:30am to 8:15pm.9. N10: 5 working days 2 days off, 10:30am to 8:15pm , 7:30am to 5:15pm.10. AA/BB/CC/A/B/C : 4 working days 2 days off, 6:30am to 6:30pm , 6:30pm to 6:30am.11. AA1/BB1/CC1/A1/B1/C1 : 4 working days 2 days off, 6:30am to 6:30pm , 6:30pm to 6:30am.12. GG/HH/II/GG1/HH1/II1 : 4 working days 2 days off, 7:30am to 7:30pm , 7:30pm to 7:30am.13. P1 : Monday to Thursday (4 working days 2 days off), 6:30am to 6:30pm , 6:30pm to 6:30am.14. P2 : Tuesday to Friday (4 working days 2 days off), 6:30am to 6:30pm , 6:30pm to 6:30am. 15. U1/U2/U3/UU1/UU2/UU3 : 4 working days 2 days off, 7:30am to 7.30pm. 16. V1/V2/V3/VV1/VV...	1
  What is the maximum allowed working hours in a week?	1. Employees are not allowed to work more than 60 hours in a week inclusive of overtime and 1 rest day per week. Company will monitor overtime and rest day utilization and take appropriate action to address instances deemed excessive.	1
  Why the company is not allowed working hours in a week more than 60 hours?	1. Continuous overtime causes worker strain that may lead to reduced productivity, increased turnover and increased injury and illnesses.	1

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 1
• per_device_eval_batch_size: 1
• num_train_epochs: 1
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: no
• prediction_loss_only: True
• per_device_train_batch_size: 1
• per_device_eval_batch_size: 1
• 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: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1
• num_train_epochs: 1
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.0
• 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: False
• 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
• 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
• dispatch_batches: None
• split_batches: 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
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin

Framework Versions

• Python: 3.11.11
• Sentence Transformers: 3.4.1
• Transformers: 4.48.2
• PyTorch: 2.5.1+cu124
• Accelerate: 1.2.1
• Datasets: 3.2.0
• Tokenizers: 0.21.0

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",
}