Upload folder using huggingface_hub

2033599 verified 5 months ago

17.1 kB

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:44114
  - loss:ContrastiveLoss
widget:
  - source_sentence: >-
      The Sadrist movement left the Alliance before the elections in December
      2005 , which also brought the Iraqi National Congress more firmly to the
      Alliance .
    sentences:
      - >-
        The Iraqi National Congress left the Alliance before the December 2005
        elections , which also brought the Sadrist movement more to the Alliance
        .
      - >-
        He pioneered important developments in the style of sculpting in wood ,
        parallel to those driven by Filippo Parodi in marble sculpture and
        Domenico Piola in painting .
      - >-
        The Mine South Deep is a large mine in the northern part of Gauteng in
        South Africa .
  - source_sentence: >-
      Mike Monroney was challenged by A.S. Thomas in the Democratic Prefix in
      1950 .
    sentences:
      - was challenged in 1950 by A.S. Mike Monroney in the Democratic Primary .
      - >-
        The T helper cells then activate the B cells , which are also in the
        presence of these antigens , causing the production of autoantibodies .
      - >-
        Illinois Route 158 , or Washington Avenue , leads west to Columbia and
        east to Belleville .
  - source_sentence: >-
      Morrow can mean either the next day in particular or the future in general
      .
    sentences:
      - >-
        Brockton is located approximately 25 miles northeast of Providence ,
        Rhode Island and 30 miles south of Boston .
      - >-
        He had been in the state playing for Melbourne , but moved to Victoria
        in 1925 and appointed New Town .
      - >-
        Morrow can either mean the next day in general , or the future in
        particular .
  - source_sentence: >-
      Fotbal Club Forex Braşov was a Romanian professional club from Braşov ,
      Romania , who was founded in October 2002 and was dissolved in 2011 .
    sentences:
      - >-
        Fotbal Club Forex Braşov was a Romanian professional club from Braşov ,
        Romania , which was dissolved in October 2002 and was founded in 2011 .
      - Nate decides to struggle for Ricky and confirms his love for her .
      - >-
        Ricardo Lingan Baccay was ordained a priest on April 10 , 1987 by
        Diosdado Aenlle Talamayan .
  - source_sentence: He was born in July 1973 in Petroupoli ( Athens ) .
    sentences:
      - >-
        Carmen Aub Romero ( born October 24 , 1989 in Mexico City , DF , Mexico
        ) is a Mexican actress .
      - He was born in Athens in July 1973 ( Petroupoli ) .
      - >-
        At the age of nine , Garcia appeared in his first concert and since then
        has appeared alone or with his aunt and his uncle in all parts of France
        .
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
  - cosine_accuracy
  - cosine_accuracy_threshold
  - cosine_f1
  - cosine_f1_threshold
  - cosine_precision
  - cosine_recall
  - cosine_ap
  - cosine_mcc
model-index:
  - name: SentenceTransformer
    results:
      - task:
          type: binary-classification
          name: Binary Classification
        dataset:
          name: paws val deberta
          type: paws-val-deberta
        metrics:
          - type: cosine_accuracy
            value: 0.9121457489878543
            name: Cosine Accuracy
          - type: cosine_accuracy_threshold
            value: 0.8481842279434204
            name: Cosine Accuracy Threshold
          - type: cosine_f1
            value: 0.9024280575539567
            name: Cosine F1
          - type: cosine_f1_threshold
            value: 0.8432618379592896
            name: Cosine F1 Threshold
          - type: cosine_precision
            value: 0.8860927152317881
            name: Cosine Precision
          - type: cosine_recall
            value: 0.9193770041227668
            name: Cosine Recall
          - type: cosine_ap
            value: 0.9503471324249102
            name: Cosine Ap
          - type: cosine_mcc
            value: 0.8230430822451054
            name: Cosine Mcc

SentenceTransformer

This is a sentence-transformers model trained. 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
Maximum Sequence Length: 64 tokens
Output Dimensionality: 768 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': 64, 'do_lower_case': False, 'architecture': '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("sentence_transformers_model_id")
# Run inference
sentences = [
    'He was born in July 1973 in Petroupoli ( Athens ) .',
    'He was born in Athens in July 1973 ( Petroupoli ) .',
    'At the age of nine , Garcia appeared in his first concert and since then has appeared alone or with his aunt and his uncle in all parts of France .',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.9386, 0.5843],
#         [0.9386, 1.0000, 0.5614],
#         [0.5843, 0.5614, 1.0000]])

Evaluation

Metrics

Binary Classification

Dataset: paws-val-deberta
Evaluated with BinaryClassificationEvaluator

Metric	Value
cosine_accuracy	0.9121
cosine_accuracy_threshold	0.8482
cosine_f1	0.9024
cosine_f1_threshold	0.8433
cosine_precision	0.8861
cosine_recall	0.9194
cosine_ap	0.9503
cosine_mcc	0.823

Training Details

Training Dataset

Unnamed Dataset

Size: 44,114 training samples
Columns: sentence_0, sentence_1, and label
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1 label
type string string float
details
min: 8 tokens
mean: 25.39 tokens
max: 47 tokens

min: 8 tokens
mean: 25.47 tokens
max: 45 tokens

min: 0.0
mean: 0.48
max: 1.0

	sentence_0	sentence_1	label
type	string	string	float
details	min: 8 tokens mean: 25.39 tokens max: 47 tokens	min: 8 tokens mean: 25.47 tokens max: 45 tokens	min: 0.0 mean: 0.48 max: 1.0

Samples:

sentence_0	sentence_1	label
`The Song of Ceylon is a 1934 British documentary film produced by Basil Wright and directed by John Grierson for the Ceylon Tea Propaganda Board .`	`The Song of Ceylon is a British documentary film directed by Basil Wright by John Grierson for the Ceylon Tea Propaganda Board in 1934 .`	`0.0`
`The two leased aircraft were returned to the BAE Systems lessor on 9 November 2006 .`	`Centavia 's two leased aircraft were returned to the lessor , BAE Systems , on November 9 , 2006 .`	`1.0`
`When , in 1818 , Ortona was assigned to Lanciano , Campli was joined to the diocese of Teramo .`	`When Ortona was assigned to Lanciano in 1818 , Campli was connected to the Diocese of Teramo .`	`1.0`

Loss: ContrastiveLoss with these parameters:

{
    "distance_metric": "SiameseDistanceMetric.COSINE_DISTANCE",
    "margin": 0.5,
    "size_average": true
}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 16
per_device_eval_batch_size: 16
num_train_epochs: 2
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 16
per_device_eval_batch_size: 16
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: 5e-05
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1
num_train_epochs: 2
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: {}
warmup_ratio: 0.0
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
bf16: False
fp16: False
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}
parallelism_config: None
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch_fused
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
project: huggingface
trackio_space_id: trackio
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
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: no
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: True
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	paws-val-deberta_cosine_ap
0.1813	500	0.0314	-
0.3626	1000	0.023	-
0.5439	1500	0.0188	-
0.7252	2000	0.0161	-
0.9065	2500	0.0148	-
1.0	2758	-	0.9361
1.0877	3000	0.0121	-
1.2690	3500	0.0107	-
1.4503	4000	0.01	-
1.6316	4500	0.0098	-
1.8129	5000	0.0094	-
1.9942	5500	0.0091	-
2.0	5516	-	0.9503

Framework Versions

Python: 3.12.12
Sentence Transformers: 5.2.0
Transformers: 4.57.3
PyTorch: 2.9.0+cu126
Accelerate: 1.12.0
Datasets: 4.0.0
Tokenizers: 0.22.1

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",
}

ContrastiveLoss

@inproceedings{hadsell2006dimensionality,
    author={Hadsell, R. and Chopra, S. and LeCun, Y.},
    booktitle={2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06)},
    title={Dimensionality Reduction by Learning an Invariant Mapping},
    year={2006},
    volume={2},
    number={},
    pages={1735-1742},
    doi={10.1109/CVPR.2006.100}
}