SentenceTransformer based on google/bert_uncased_L-2_H-128_A-2

This is a sentence-transformers model finetuned from google/bert_uncased_L-2_H-128_A-2. It maps sentences & paragraphs to a 128-dimensional dense vector space and can be used for retrieval.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: google/bert_uncased_L-2_H-128_A-2
• Maximum Sequence Length: 128 tokens
• Output Dimensionality: 128 dimensions
• Similarity Function: Cosine Similarity
• Supported Modality: Text

Model Sources

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

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'transformer_task': 'feature-extraction', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'last_hidden_state'}}, 'module_output_name': 'token_embeddings', 'architecture': 'BertModel'})
  (1): Pooling({'embedding_dimension': 128, 'pooling_mode': 'mean', 'include_prompt': True})
)

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("swardiantara/bert-tiny-amazon_reviews-k3-fixed-euclidean")
# Run inference
sentences = [
    'Only worked for about a month and now dead?',
    "The blades have come apart. Very hard to use. I guess it's more for commercial use and bolted to a table.",
    "I don't know - for some reason I was disappointed in this product. It was as advertised but the pictures are so small and not put into a work out routine - seems scattered. Would have returned them but repackaging them would have been a pain.",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 128]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.9976, 0.9982],
#         [0.9976, 1.0000, 0.9971],
#         [0.9982, 0.9971, 1.0000]])

Training Details

Training Dataset

Unnamed Dataset

• Size: 2,600,030 training samples
• Columns: text_a, text_b, and label
• Approximate statistics based on the first 100 samples:

  	text_a	text_b	label
  type	string	string	list
  modality	text	text
  details	min: 11 tokensmean: 41.25 tokensmax: 128 tokens	min: 17 tokensmean: 52.32 tokensmax: 101 tokens	size: 2 elements

• Samples:

  text_a	text_b	label
  Arrived broken. Manufacturer defect. Two of the legs of the base were not completely formed, so there was no way to insert the casters. I unpackaged the entire chair and hardware before noticing this. So, I'll spend twice the amount of time boxing up the whole useless thing and send it back with a 1-star review of part of a chair I never got to sit in. I will go so far as to include a picture of what their injection molding and quality assurance process missed though. I will be hesitant to buy again. It makes me wonder if there aren't missing structures and supports that don't impede the assembly process.	I ordered two of these. One came broken. It was packaged okay, so I believe it was broken due to the mail carrier tossing it onto my porch. The one that did not break, I love. It looks nice on my counter and serves its purpose. Updated review: Ok so the salt cellar that did not break during shipping pretty much shattered at the lightest touch after 3 weeks on my counter. I can no longer recommend this item, clearly they are not very durable at all.	[1.0, 0.0]
  Arrived broken. Manufacturer defect. Two of the legs of the base were not completely formed, so there was no way to insert the casters. I unpackaged the entire chair and hardware before noticing this. So, I'll spend twice the amount of time boxing up the whole useless thing and send it back with a 1-star review of part of a chair I never got to sit in. I will go so far as to include a picture of what their injection molding and quality assurance process missed though. I will be hesitant to buy again. It makes me wonder if there aren't missing structures and supports that don't impede the assembly process.	Seems to protect well, feels nice, just wish it would stick a bit better. The first one would not stay applied. Kept peeling from the top no matter what I did. I figured it could be my error, so I removed and cleaned the screen very well with additional alcohol wipes. The second one has been on a few days, but is starting to peel from a different spot. I love the feel of the protector otherwise.	[0.0, 0.25]
  Arrived broken. Manufacturer defect. Two of the legs of the base were not completely formed, so there was no way to insert the casters. I unpackaged the entire chair and hardware before noticing this. So, I'll spend twice the amount of time boxing up the whole useless thing and send it back with a 1-star review of part of a chair I never got to sit in. I will go so far as to include a picture of what their injection molding and quality assurance process missed though. I will be hesitant to buy again. It makes me wonder if there aren't missing structures and supports that don't impede the assembly process.	I ended up returning this bag for credit. While there was nothing wrong with the quality, it turned out to be too small.	[0.0, 0.25]

• Loss: main.OrdinalProxyContrastiveLoss

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 1024
• learning_rate: 1e-05
• load_best_model_at_end: True

All Hyperparameters

• per_device_train_batch_size: 1024
• num_train_epochs: 3
• max_steps: -1
• learning_rate: 1e-05
• lr_scheduler_type: linear
• lr_scheduler_kwargs: None
• warmup_steps: 0
• optim: adamw_torch
• optim_args: None
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• optim_target_modules: None
• gradient_accumulation_steps: 1
• average_tokens_across_devices: True
• max_grad_norm: 1.0
• label_smoothing_factor: 0.0
• bf16: False
• fp16: False
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• use_liger_kernel: False
• liger_kernel_config: None
• use_cache: False
• neftune_noise_alpha: None
• torch_empty_cache_steps: None
• auto_find_batch_size: False
• log_on_each_node: True
• logging_nan_inf_filter: True
• include_num_input_tokens_seen: no
• log_level: passive
• log_level_replica: warning
• disable_tqdm: False
• project: huggingface
• trackio_space_id: None
• trackio_bucket_id: None
• trackio_static_space_id: None
• per_device_eval_batch_size: 8
• prediction_loss_only: True
• eval_on_start: False
• eval_do_concat_batches: True
• eval_use_gather_object: False
• eval_accumulation_steps: None
• include_for_metrics: []
• batch_eval_metrics: False
• save_only_model: False
• save_on_each_node: False
• enable_jit_checkpoint: False
• push_to_hub: False
• hub_private_repo: None
• hub_model_id: None
• hub_strategy: every_save
• hub_always_push: False
• hub_revision: None
• load_best_model_at_end: True
• ignore_data_skip: False
• restore_callback_states_from_checkpoint: False
• full_determinism: False
• seed: 42
• data_seed: None
• use_cpu: False
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• parallelism_config: None
• dataloader_drop_last: False
• dataloader_num_workers: 0
• dataloader_pin_memory: True
• dataloader_persistent_workers: False
• dataloader_prefetch_factor: None
• remove_unused_columns: True
• label_names: None
• train_sampling_strategy: random
• length_column_name: length
• ddp_find_unused_parameters: None
• ddp_bucket_cap_mb: None
• ddp_broadcast_buffers: False
• ddp_static_graph: None
• ddp_backend: None
• ddp_timeout: 1800
• fsdp: None
• fsdp_config: None
• deepspeed: None
• debug: []
• skip_memory_metrics: True
• do_predict: False
• resume_from_checkpoint: None
• warmup_ratio: None
• local_rank: -1
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

• The bold row denotes the saved checkpoint.

Training Time

• Training: 26.6 minutes
• Evaluation: 1.3 seconds
• Total: 26.6 minutes

Framework Versions

• Python: 3.12.4
• Sentence Transformers: 5.5.1
• Transformers: 5.11.0
• PyTorch: 2.5.1+cu121
• Accelerate: 1.13.0
• Datasets: 2.21.0
• Tokenizers: 0.22.2

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",
}