bobox/DeBERTaV3-xSmall-SentenceTransformer-0.03

SentenceTransformer based on microsoft/deberta-v3-xsmall

This is a sentence-transformers model finetuned from microsoft/deberta-v3-xsmall on the stanfordnlp/snli 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: microsoft/deberta-v3-xsmall
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 384 tokens
• Similarity Function: Cosine Similarity
• Training Dataset:
  • stanfordnlp/snli
• Language: en

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}) with Transformer model: DebertaV2Model 
  (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})
)

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("bobox/DeBERTaV3-xSmall-SentenceTransformer-0.03")
# Run inference
sentences = [
    'in each square',
    'It is widespread.',
    'A young girl flips an omelet.',
]
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]

Evaluation

Metrics

Semantic Similarity

• Dataset: sts-dev
• Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.7972
spearman_cosine	0.807
pearson_manhattan	0.8079
spearman_manhattan	0.8072
pearson_euclidean	0.8084
spearman_euclidean	0.8073
pearson_dot	0.7029
spearman_dot	0.6909
pearson_max	0.8084
spearman_max	0.8073

Binary Classification

• Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.6772
cosine_accuracy_threshold	0.7285
cosine_f1	0.7187
cosine_f1_threshold	0.6111
cosine_precision	0.611
cosine_recall	0.8724
cosine_ap	0.7392
dot_accuracy	0.6383
dot_accuracy_threshold	228.4041
dot_f1	0.7068
dot_f1_threshold	177.3942
dot_precision	0.5811
dot_recall	0.9017
dot_ap	0.6904
manhattan_accuracy	0.6635
manhattan_accuracy_threshold	174.6275
manhattan_f1	0.7054
manhattan_f1_threshold	232.6788
manhattan_precision	0.5772
manhattan_recall	0.907
manhattan_ap	0.7282
euclidean_accuracy	0.6651
euclidean_accuracy_threshold	13.4225
euclidean_f1	0.7068
euclidean_f1_threshold	17.6348
euclidean_precision	0.5756
euclidean_recall	0.9154
euclidean_ap	0.7303
max_accuracy	0.6772
max_accuracy_threshold	228.4041
max_f1	0.7187
max_f1_threshold	232.6788
max_precision	0.611
max_recall	0.9154
max_ap	0.7392

Training Details

Training Dataset

stanfordnlp/snli

• Dataset: stanfordnlp/snli at cdb5c3d
• Size: 314,315 training samples
• Columns: sentence1, sentence2, and label
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	label
  type	string	string	int
  details	min: 5 tokens mean: 16.62 tokens max: 62 tokens	min: 4 tokens mean: 9.46 tokens max: 29 tokens	0: 100.00%

• Samples:

  sentence1	sentence2	label
  A person on a horse jumps over a broken down airplane.	A person is outdoors, on a horse.	0
  Children smiling and waving at camera	There are children present	0
  A boy is jumping on skateboard in the middle of a red bridge.	The boy does a skateboarding trick.	0

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Evaluation Dataset

sentence-transformers/stsb

• Dataset: sentence-transformers/stsb at ab7a5ac
• Size: 1,500 evaluation samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 5 tokens mean: 14.77 tokens max: 45 tokens	min: 6 tokens mean: 14.74 tokens max: 49 tokens	min: 0.0 mean: 0.47 max: 1.0

• Samples:

  sentence1	sentence2	score
  A man with a hard hat is dancing.	A man wearing a hard hat is dancing.	1.0
  A young child is riding a horse.	A child is riding a horse.	0.95
  A man is feeding a mouse to a snake.	The man is feeding a mouse to the snake.	1.0

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• learning_rate: 7.5e-05
• num_train_epochs: 2
• warmup_ratio: 0.25
• save_safetensors: False
• fp16: True
• push_to_hub: True
• hub_model_id: bobox/DeBERTaV3-xSmall-SentenceTransformer-0.03n
• hub_strategy: checkpoint

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
• learning_rate: 7.5e-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: 2
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.25
• warmup_steps: 0
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• save_safetensors: False
• 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: True
• resume_from_checkpoint: None
• hub_model_id: bobox/DeBERTaV3-xSmall-SentenceTransformer-0.03n
• hub_strategy: checkpoint
• hub_private_repo: False
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• 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
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss	max_ap	sts-dev_spearman_cosine
None	0	-	3.7624	0.5721	0.4168
0.0501	246	3.3825	-	-	-
0.1002	492	1.8307	-	-	-
0.1500	737	-	1.0084	0.7024	-
0.1502	738	1.055	-	-	-
0.2003	984	0.7961	-	-	-
0.2504	1230	0.6859	-	-	-
0.3001	1474	-	0.7410	0.7191	-
0.3005	1476	0.5914	-	-	-
0.3506	1722	0.5324	-	-	-
0.4007	1968	0.5077	-	-	-
0.4501	2211	-	0.6152	0.7144	-
0.4507	2214	0.4647	-	-	-
0.5008	2460	0.4443	-	-	-
0.5509	2706	0.4169	-	-	-
0.6002	2948	-	0.5820	0.7207	-
0.6010	2952	0.3831	-	-	-
0.6511	3198	0.393	-	-	-
0.7011	3444	0.3654	-	-	-
0.7502	3685	-	0.5284	0.7264	-
0.7512	3690	0.344	-	-	-
0.8013	3936	0.3336	-	-	-
0.8514	4182	0.3382	-	-	-
0.9002	4422	-	0.4911	0.7294	-
0.9015	4428	0.3182	-	-	-
0.9515	4674	0.3213	-	-	-
1.0016	4920	0.3032	-	-	-
1.0503	5159	-	0.4777	0.7325	-
1.0517	5166	0.2526	-	-	-
1.1018	5412	0.2652	-	-	-
1.1519	5658	0.2538	-	-	-
1.2003	5896	-	0.4569	0.7331	-
1.2020	5904	0.2454	-	-	-
1.2520	6150	0.2528	-	-	-
1.3021	6396	0.2448	-	-	-
1.3504	6633	-	0.4334	0.7370	-
1.3522	6642	0.2282	-	-	-
1.4023	6888	0.2295	-	-	-
1.4524	7134	0.2313	-	-	-
1.5004	7370	-	0.4237	0.7342	-
1.5024	7380	0.2218	-	-	-
1.5525	7626	0.2246	-	-	-
1.6026	7872	0.218	-	-	-
1.6504	8107	-	0.4102	0.7388	-
1.6527	8118	0.2095	-	-	-
1.7028	8364	0.2114	-	-	-
1.7529	8610	0.2063	-	-	-
1.8005	8844	-	0.4075	0.7370	-
1.8029	8856	0.1968	-	-	-
1.8530	9102	0.2061	-	-	-
1.9031	9348	0.2089	-	-	-
1.9505	9581	-	0.3978	0.7395	-
1.9532	9594	0.2005	-	-	-
2.0	9824	-	0.3963	0.7392	-
None	0	-	1.5506	-	0.8070

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 3.0.0
• Transformers: 4.41.2
• PyTorch: 2.3.0+cu121
• Accelerate: 0.30.1
• Datasets: 2.19.2
• Tokenizers: 0.19.1

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",
}

MultipleNegativesRankingLoss

@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply}, 
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}