metadata

language: []
library_name: sentence-transformers
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:38688
  - loss:ContrastiveLoss
base_model: sentence-transformers/all-MiniLM-L6-v2
datasets: []
widget:
  - source_sentence: >-
      There is a heavy cost for this service provided in conjunction with NOAA
      and SARSAT.
    sentences:
      - >-
        No significant changes have been made to the roadway except for its
        legal definition.
      - Some academics have questioned the ethics of these payments.
      - >-
        There is no charge for this service provided in conjunction with NOAA
        and SARSAT.
  - source_sentence: You're not thin.
    sentences:
      - This process is called low-dimensional embedded in machine learning.
      - You're thin.
      - Jean Prouvost was the founder of Marie Claire.
  - source_sentence: The lead man is charisma-free.
    sentences:
      - >-
        Fossil egg s are rare, but one oogenus, Polyclonoolithus, was discovered
        in the Hekou Group.
      - The roof is shingled, and topped by a small belfry.
      - The lead man doesn't have charisma.
  - source_sentence: >-
      Willis has criticized the rules adopted by the RNC, particularly Rules 12,
      16, and 40.
    sentences:
      - >-
        Willis has fully accepted the rules adopted by the RNC, particularly
        Rules 12, 16, and 40.
      - I can't stop reading.
      - This force acts on water independently of the wind stress.
  - source_sentence: The publication was named after Sir James Joynton Smith.
    sentences:
      - >-
        Detailed specific information on the ongoing validation activities is
        being made available in related publications.
      - On November 25, 2012, Tom O'Brien was reinstated.
      - >-
        The publication took its name from its founder and chief financer Sir
        James Joynton Smith.
pipeline_tag: sentence-similarity

SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L6-v2. 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: sentence-transformers/all-MiniLM-L6-v2
Maximum Sequence Length: 256 tokens
Output Dimensionality: 384 tokens
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': 256, 'do_lower_case': False}) with Transformer model: BertModel 
  (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})
  (2): Normalize()
)

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("LeoChiuu/all-MiniLM-L6-v2-negations")
# Run inference
sentences = [
    'The publication was named after Sir James Joynton Smith.',
    'The publication took its name from its founder and chief financer Sir James Joynton Smith.',
    "On November 25, 2012, Tom O'Brien was reinstated.",
]
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]

Training Details

Training Dataset

Unnamed Dataset

Size: 38,688 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 int
details
min: 5 tokens
mean: 15.94 tokens
max: 41 tokens

min: 5 tokens
mean: 15.96 tokens
max: 44 tokens

0: ~48.50%
1: ~51.50%

	sentence_0	sentence_1	label
type	string	string	int
details	min: 5 tokens mean: 15.94 tokens max: 41 tokens	min: 5 tokens mean: 15.96 tokens max: 44 tokens	0: ~48.50% 1: ~51.50%

Samples:

sentence_0	sentence_1	label
`No, that is impossible.`	`No, that is not possible.`	`0`
`The building did indeed serve as a hof, according to the bone finds.`	`The bone finds thus indicate the building did indeed serve as a hof.`	`0`
`The building became a pet shop.`	`The building became a hospital.`	`1`

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: 10
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
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
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: 10
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
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: 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, '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: 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_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	Training Loss
0.2068	500	0.0353
0.4136	1000	0.0307
0.6203	1500	0.0234
0.8271	2000	0.0187
1.0339	2500	0.0152
1.2407	3000	0.0134
1.4475	3500	0.0123
1.6543	4000	0.0111
1.8610	4500	0.0107
2.0678	5000	0.0097
2.2746	5500	0.0096
2.4814	6000	0.0091
2.6882	6500	0.0087
2.8950	7000	0.0086
3.1017	7500	0.0075
3.3085	8000	0.008
3.5153	8500	0.0074
3.7221	9000	0.007
3.9289	9500	0.007
4.1356	10000	0.0063
4.3424	10500	0.0068
4.5492	11000	0.0061
4.7560	11500	0.0059
4.9628	12000	0.0056
5.1696	12500	0.0052
5.3763	13000	0.0055
5.5831	13500	0.0051
5.7899	14000	0.005
5.9967	14500	0.0047
6.2035	15000	0.0046
6.4103	15500	0.0047
6.6170	16000	0.0044
6.8238	16500	0.0044
7.0306	17000	0.0041
7.2374	17500	0.004
7.4442	18000	0.0044
7.6510	18500	0.0039
7.8577	19000	0.0038
8.0645	19500	0.0038
8.2713	20000	0.0037
8.4781	20500	0.0039
8.6849	21000	0.0037
8.8916	21500	0.0036
9.0984	22000	0.0034
9.3052	22500	0.0036
9.5120	23000	0.0035
9.7188	23500	0.0034
9.9256	24000	0.0035

Framework Versions

Python: 3.11.9
Sentence Transformers: 3.0.1
Transformers: 4.40.2
PyTorch: 2.3.0+cpu
Accelerate: 0.32.1
Datasets: 2.19.1
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",
}

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}
}