epsil/TinyBERT_L-4_H-312_v2-distilled-from-stsb-roberta-base-v2

SentenceTransformer based on nreimers/TinyBERT_L-4_H-312_v2

This is a sentence-transformers model finetuned from nreimers/TinyBERT_L-4_H-312_v2. It maps sentences & paragraphs to a 312-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: nreimers/TinyBERT_L-4_H-312_v2
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 312 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': 512, 'do_lower_case': False, 'architecture': 'BertModel'})
  (1): Pooling({'word_embedding_dimension': 312, '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})
)

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("epsil/TinyBERT_L-4_H-312_v2-distilled-from-stsb-roberta-base-v2")
# Run inference
sentences = [
    'A black dog is drinking next to a brown and white dog that is looking at an orange ball in the lake, whilst a horse and rider passes behind.',
    'A man with a white towel wrapped around the lower part of his face and neck.',
    'There are two people running around a track in lane three and the one wearing a blue shirt with a green thing over the eyes is just barely ahead of the guy wearing an orange shirt and sunglasses.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 312]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[ 1.0000, -0.1689,  0.1228],
#         [-0.1689,  1.0000,  0.0546],
#         [ 0.1228,  0.0546,  1.0000]])

Evaluation

Metrics

Semantic Similarity

• Datasets: sts-dev and sts-test
• Evaluated with EmbeddingSimilarityEvaluator

Metric	sts-dev	sts-test
pearson_cosine	0.8038	0.7516
spearman_cosine	0.8178	0.7563

Knowledge Distillation

• Evaluated with MSEEvaluator

Metric	Value
negative_mse	-50.0177

Training Details

Training Dataset

Unnamed Dataset

• Size: 200,000 training samples
• Columns: sentence and label
• Approximate statistics based on the first 1000 samples:

  	sentence	label
  type	string	list
  details	min: 4 tokens mean: 12.24 tokens max: 52 tokens	size: 312 elements

• Samples:

  sentence	label
  A person on a horse jumps over a broken down airplane.	[0.07039763033390045, 0.7007468938827515, -2.6371383666992188, 1.7311089038848877, 1.122781753540039, ...]
  Children smiling and waving at camera	[-2.568326711654663, 3.1153242588043213, 7.387216091156006, 5.154618263244629, -2.5198936462402344, ...]
  A boy is jumping on skateboard in the middle of a red bridge.	[3.0327019691467285, 2.922370433807373, 1.2597863674163818, 6.1974382400512695, -0.8628579378128052, ...]

• Loss: MSELoss

Evaluation Dataset

Unnamed Dataset

• Size: 10,000 evaluation samples
• Columns: sentence and label
• Approximate statistics based on the first 1000 samples:

  	sentence	label
  type	string	list
  details	min: 5 tokens mean: 13.23 tokens max: 57 tokens	size: 312 elements

• Samples:

  sentence	label
  Two women are embracing while holding to go packages.	[-6.152304172515869, -1.9879305362701416, 2.1665844917297363, -2.0057384967803955, 1.4534344673156738, ...]
  Two young children in blue jerseys, one with the number 9 and one with the number 2 are standing on wooden steps in a bathroom and washing their hands in a sink.	[-1.7411372661590576, 0.6246002912521362, 2.5846199989318848, 3.96124267578125, -2.789034843444824, ...]
  A man selling donuts to a customer during a world exhibition event held in the city of Angeles	[3.279698371887207, 3.120692253112793, -0.29934388399124146, -2.4101784229278564, 3.1145691871643066, ...]

• Loss: MSELoss

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• learning_rate: 0.0001
• num_train_epochs: 1
• warmup_ratio: 0.1
• fp16: True
• load_best_model_at_end: True

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• 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: 0.0001
• 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: True
• 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: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	Validation Loss	sts-dev_spearman_cosine	negative_mse	sts-test_spearman_cosine
0.032	100	0.8834	-	-	-	-
0.064	200	0.8003	-	-	-	-
0.096	300	0.6854	-	-	-	-
0.128	400	0.6016	-	-	-	-
0.16	500	0.5553	0.6273	0.7637	-62.7347	-
0.192	600	0.523	-	-	-	-
0.224	700	0.4987	-	-	-	-
0.256	800	0.482	-	-	-	-
0.288	900	0.4627	-	-	-	-
0.32	1000	0.4477	0.5635	0.7950	-56.3465	-
0.352	1100	0.4351	-	-	-	-
0.384	1200	0.4251	-	-	-	-
0.416	1300	0.4151	-	-	-	-
0.448	1400	0.4077	-	-	-	-
0.48	1500	0.403	0.5329	0.8085	-53.2905	-
0.512	1600	0.3905	-	-	-	-
0.544	1700	0.3883	-	-	-	-
0.576	1800	0.3825	-	-	-	-
0.608	1900	0.3761	-	-	-	-
0.64	2000	0.3721	0.5145	0.8133	-51.4495	-
0.672	2100	0.3696	-	-	-	-
0.704	2200	0.3674	-	-	-	-
0.736	2300	0.3644	-	-	-	-
0.768	2400	0.3597	-	-	-	-
0.8	2500	0.3558	0.5052	0.8161	-50.5228	-
0.832	2600	0.3524	-	-	-	-
0.864	2700	0.3521	-	-	-	-
0.896	2800	0.3504	-	-	-	-
0.928	2900	0.3499	-	-	-	-
0.96	3000	0.35	0.5002	0.8178	-50.0177	-
0.992	3100	0.348	-	-	-	-
-1	-1	-	-	-	-	0.7563

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.12.3
• Sentence Transformers: 5.2.0
• Transformers: 4.49.0
• PyTorch: 2.4.1+cu121
• Accelerate: 1.12.0
• Datasets: 3.0.1
• 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",
}

MSELoss

@inproceedings{reimers-2020-multilingual-sentence-bert,
    title = "Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillation",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2020",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/2004.09813",
}