bobox/DeBERTaV3-small-ST-AdaptiveLayers-ep2

SentenceTransformer based on microsoft/deberta-v3-small

This is a sentence-transformers model finetuned from microsoft/deberta-v3-small on the stanfordnlp/snli dataset. It maps sentences & paragraphs to a 768-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-small
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 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': 768, '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-small-ST-AdaptiveLayers-ep2")
# Run inference
sentences = [
    'A wet child stands in chest deep ocean water.',
    'The child s playing on the beach.',
    'A woman paints a portrait of her best friend.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

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

Evaluation

Metrics

Binary Classification

• Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.6583
cosine_accuracy_threshold	0.6767
cosine_f1	0.7049
cosine_f1_threshold	0.6018
cosine_precision	0.6115
cosine_recall	0.8321
cosine_ap	0.6995
dot_accuracy	0.6272
dot_accuracy_threshold	163.2505
dot_f1	0.6976
dot_f1_threshold	119.2078
dot_precision	0.5639
dot_recall	0.9144
dot_ap	0.6437
manhattan_accuracy	0.6571
manhattan_accuracy_threshold	243.7545
manhattan_f1	0.7056
manhattan_f1_threshold	295.9595
manhattan_precision	0.5901
manhattan_recall	0.8773
manhattan_ap	0.7072
euclidean_accuracy	0.6591
euclidean_accuracy_threshold	12.1418
euclidean_f1	0.7037
euclidean_f1_threshold	14.1975
euclidean_precision	0.5997
euclidean_recall	0.8513
euclidean_ap	0.7035
max_accuracy	0.6591
max_accuracy_threshold	243.7545
max_f1	0.7056
max_f1_threshold	295.9595
max_precision	0.6115
max_recall	0.9144
max_ap	0.7072

Semantic Similarity

• Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.7322
spearman_cosine	0.7345
pearson_manhattan	0.7537
spearman_manhattan	0.7551
pearson_euclidean	0.7468
spearman_euclidean	0.7485
pearson_dot	0.6143
spearman_dot	0.61
pearson_max	0.7537
spearman_max	0.7551

Training Details

Training Dataset

stanfordnlp/snli

• Dataset: stanfordnlp/snli at cdb5c3d
• Size: 67,190 training samples
• Columns: sentence1, sentence2, and label
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	label
  type	string	string	int
  details	min: 4 tokens mean: 21.19 tokens max: 133 tokens	min: 4 tokens mean: 11.77 tokens max: 49 tokens	0: 100.00%

• Samples:

  sentence1	sentence2	label
  Without a placebo group, we still won't know if any of the treatments are better than nothing and therefore worth giving.	It is necessary to use a controlled method to ensure the treatments are worthwhile.	0
  It was conducted in silence.	It was done silently.	0
  oh Lewisville any decent food in your cafeteria up there	Is there any decent food in your cafeteria up there in Lewisville?	0

• Loss: AdaptiveLayerLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "n_layers_per_step": 1,
      "last_layer_weight": 1,
      "prior_layers_weight": 1,
      "kl_div_weight": 1,
      "kl_temperature": 1
  }
  

Evaluation Dataset

stanfordnlp/snli

• Dataset: stanfordnlp/snli at cdb5c3d
• 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: AdaptiveLayerLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "n_layers_per_step": 1,
      "last_layer_weight": 1,
      "prior_layers_weight": 1,
      "kl_div_weight": 1,
      "kl_temperature": 1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 42
• per_device_eval_batch_size: 22
• learning_rate: 3e-06
• weight_decay: 1e-08
• num_train_epochs: 2
• lr_scheduler_type: cosine
• warmup_ratio: 0.5
• save_safetensors: False
• fp16: True
• hub_model_id: bobox/DeBERTaV3-small-ST-AdaptiveLayers-ep2-tmp
• hub_strategy: checkpoint
• hub_private_repo: True
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 42
• per_device_eval_batch_size: 22
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• learning_rate: 3e-06
• weight_decay: 1e-08
• 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: cosine
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.5
• 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: False
• resume_from_checkpoint: None
• hub_model_id: bobox/DeBERTaV3-small-ST-AdaptiveLayers-ep2-tmp
• hub_strategy: checkpoint
• hub_private_repo: True
• 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: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss	max_ap	spearman_cosine
0.1	160	4.6003	4.8299	0.6017	-
0.2	320	4.0659	4.3436	0.6168	-
0.3	480	3.4886	4.0840	0.6339	-
0.4	640	3.0592	3.6422	0.6611	-
0.5	800	2.5728	3.1927	0.6773	-
0.6	960	2.184	2.8322	0.6893	-
0.7	1120	1.8744	2.4892	0.6954	-
0.8	1280	1.757	2.4453	0.7002	-
0.9	1440	1.5872	2.2565	0.7010	-
1.0	1600	1.446	2.1391	0.7046	-
1.1	1760	1.3892	2.1236	0.7058	-
1.2	1920	1.2567	1.9738	0.7053	-
1.3	2080	1.2233	1.8925	0.7063	-
1.4	2240	1.1954	1.8392	0.7075	-
1.5	2400	1.1395	1.9081	0.7065	-
1.6	2560	1.1211	1.8080	0.7074	-
1.7	2720	1.0825	1.8408	0.7073	-
1.8	2880	1.1358	1.7363	0.7073	-
1.9	3040	1.0628	1.8936	0.7072	-
2.0	3200	1.1412	1.7846	0.7072	-
None	0	-	3.0121	0.7072	0.7345

Framework Versions

• Python: 3.10.13
• Sentence Transformers: 3.0.1
• Transformers: 4.41.2
• PyTorch: 2.1.2
• 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",
}

AdaptiveLayerLoss

@misc{li20242d,
    title={2D Matryoshka Sentence Embeddings}, 
    author={Xianming Li and Zongxi Li and Jing Li and Haoran Xie and Qing Li},
    year={2024},
    eprint={2402.14776},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

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}
}