LamaDiab/MiniLM-V22Data-256ConstantBATCH-SemanticEngine

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, '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("LamaDiab/MiniLM-V22Data-256ConstantBATCH-SemanticEngine")
# Run inference
sentences = [
    'must kindergarten backpack mermazing 2 cases',
    'school supplies',
    'crescent stand with 3 dates plate gold',
]
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,  0.5516, -0.1659],
#         [ 0.5516,  1.0000,  0.0134],
#         [-0.1659,  0.0134,  1.0000]])

Evaluation

Metrics

Triplet

• Evaluated with TripletEvaluator

Metric	Value
cosine_accuracy	0.9688

Training Details

Training Dataset

Unnamed Dataset

• Size: 647,236 training samples
• Columns: anchor, positive, and itemCategory
• Approximate statistics based on the first 1000 samples:

  	anchor	positive	itemCategory
  type	string	string	string
  details	min: 3 tokens mean: 11.56 tokens max: 50 tokens	min: 3 tokens mean: 4.55 tokens max: 12 tokens	min: 3 tokens mean: 3.91 tokens max: 9 tokens

• Samples:

  anchor	positive	itemCategory
  petrol samsung galaxy	smart phone	smart phone
  must trolley bag must true football 4 cases	wheels cover backpack	bag
  sanpellegrino chino is a bold and refreshing italian beverage with a unique bittersweet flavor made from herbal extracts and citrus best served chilled for a distinctive taste experience	chino can drink	beverage

• Loss: MultipleNegativesSymmetricRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "gather_across_devices": false
  }
  

Evaluation Dataset

Unnamed Dataset

• Size: 9,509 evaluation samples
• Columns: anchor, positive, negative, and itemCategory
• Approximate statistics based on the first 1000 samples:

  	anchor	positive	negative	itemCategory
  type	string	string	string	string
  details	min: 3 tokens mean: 9.63 tokens max: 43 tokens	min: 3 tokens mean: 6.61 tokens max: 150 tokens	min: 3 tokens mean: 9.58 tokens max: 46 tokens	min: 3 tokens mean: 3.88 tokens max: 10 tokens

• Samples:

  anchor	positive	negative	itemCategory
  pilot mechanical pencil progrex h-127 - 0.7 mm	pencil	artist pen brush tip 1.5m gold no.250	pencil
  superior drawing marker -pen - set of 12 colors - 2 nib	superior	notte 11-101 a5 stapled squared notebook, 60 sheets, cardboard cover, 60 grams, 148 x 210 mm, turkish	marker
  first person singular author: haruki murakami	haruki murakami book	yellow dinosaur assembling game	literature and fiction

• Loss: MultipleNegativesSymmetricRankingLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "gather_across_devices": false
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 256
• per_device_eval_batch_size: 256
• learning_rate: 2e-05
• weight_decay: 0.0001
• num_train_epochs: 8
• warmup_ratio: 0.1
• fp16: True
• dataloader_num_workers: 1
• dataloader_prefetch_factor: 2
• dataloader_persistent_workers: True
• push_to_hub: True
• hub_model_id: MiniLM-V22Data-256ConstantBATCH-SemanticEngine
• hub_strategy: all_checkpoints

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 256
• per_device_eval_batch_size: 256
• 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: 0.0001
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 8
• 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: 1
• dataloader_prefetch_factor: 2
• 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: True
• skip_memory_metrics: True
• use_legacy_prediction_loop: False
• push_to_hub: True
• resume_from_checkpoint: None
• hub_model_id: MiniLM-V22Data-256ConstantBATCH-SemanticEngine
• hub_strategy: all_checkpoints
• 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

Epoch	Step	Training Loss	Validation Loss	cosine_accuracy
0.0004	1	4.3129	-	-
0.3954	1000	3.298	0.5520	0.9471
0.7908	2000	1.9673	0.4995	0.9539
1.1861	3000	1.5646	0.4841	0.9593
1.5812	4000	1.5836	0.4644	0.9627
1.9763	5000	1.4401	0.4496	0.9638
2.3714	6000	1.2966	0.4553	0.9672
2.7665	7000	1.2287	0.4436	0.9656
3.1616	8000	1.1559	0.4434	0.9663
3.5567	9000	1.1011	0.4327	0.9674
3.9518	10000	1.0585	0.4340	0.9675
4.3469	11000	1.0006	0.4300	0.9685
4.7420	12000	0.9839	0.4264	0.9687
5.1371	13000	0.9693	0.4285	0.9685
5.5322	14000	0.9279	0.4340	0.9680
5.9273	15000	0.9175	0.4266	0.9694
6.3224	16000	0.8994	0.4258	0.9680
6.7175	17000	0.8934	0.4268	0.9685
7.1126	18000	0.8777	0.4271	0.9686
7.5077	19000	0.8602	0.4265	0.9689
7.9028	20000	0.8681	0.4262	0.9688

Framework Versions

• Python: 3.11.13
• Sentence Transformers: 5.1.2
• Transformers: 4.53.3
• PyTorch: 2.6.0+cu124
• Accelerate: 1.9.0
• Datasets: 4.4.1
• Tokenizers: 0.21.2

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",
}