metadata

language: []
library_name: sentence-transformers
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:844
  - loss:CoSENTLoss
base_model: sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
datasets: []
metrics:
  - pearson_cosine
  - spearman_cosine
  - pearson_manhattan
  - spearman_manhattan
  - pearson_euclidean
  - spearman_euclidean
  - pearson_dot
  - spearman_dot
  - pearson_max
  - spearman_max
widget:
  - source_sentence: Help fix a problem with my device’s battery life
    sentences:
      - order query
      - faq query
      - technical support query
  - source_sentence: 订购一双运动鞋
    sentences:
      - service request
      - feedback query
      - product query
  - source_sentence: 告诉我如何更改我的密码
    sentences:
      - support query
      - product query
      - faq query
  - source_sentence: Get information on the next local festival
    sentences:
      - event inquiry
      - service request
      - account query
  - source_sentence: Change the currency for my payment
    sentences:
      - product query
      - payment query
      - faq query
pipeline_tag: sentence-similarity
model-index:
  - name: >-
      SentenceTransformer based on
      sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
    results:
      - task:
          type: semantic-similarity
          name: Semantic Similarity
        dataset:
          name: MiniLM dev
          type: MiniLM-dev
        metrics:
          - type: pearson_cosine
            value: 0.7356955662825808
            name: Pearson Cosine
          - type: spearman_cosine
            value: 0.7320761390174187
            name: Spearman Cosine
          - type: pearson_manhattan
            value: 0.6240041985776243
            name: Pearson Manhattan
          - type: spearman_manhattan
            value: 0.6179783414452009
            name: Spearman Manhattan
          - type: pearson_euclidean
            value: 0.6321466982201008
            name: Pearson Euclidean
          - type: spearman_euclidean
            value: 0.6296964936282937
            name: Spearman Euclidean
          - type: pearson_dot
            value: 0.7491168439451736
            name: Pearson Dot
          - type: spearman_dot
            value: 0.7592129124940543
            name: Spearman Dot
          - type: pearson_max
            value: 0.7491168439451736
            name: Pearson Max
          - type: spearman_max
            value: 0.7592129124940543
            name: Spearman Max
      - task:
          type: semantic-similarity
          name: Semantic Similarity
        dataset:
          name: MiniLM test
          type: MiniLM-test
        metrics:
          - type: pearson_cosine
            value: 0.7687106130417081
            name: Pearson Cosine
          - type: spearman_cosine
            value: 0.7552108666502075
            name: Spearman Cosine
          - type: pearson_manhattan
            value: 0.7462708006775693
            name: Pearson Manhattan
          - type: spearman_manhattan
            value: 0.7365483246407295
            name: Spearman Manhattan
          - type: pearson_euclidean
            value: 0.7545194410402545
            name: Pearson Euclidean
          - type: spearman_euclidean
            value: 0.7465016803791179
            name: Spearman Euclidean
          - type: pearson_dot
            value: 0.7251488155932073
            name: Pearson Dot
          - type: spearman_dot
            value: 0.7390366635753267
            name: Spearman Dot
          - type: pearson_max
            value: 0.7687106130417081
            name: Pearson Max
          - type: spearman_max
            value: 0.7552108666502075
            name: Spearman Max

SentenceTransformer based on sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2

This is a sentence-transformers model finetuned from sentence-transformers/paraphrase-multilingual-MiniLM-L12-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/paraphrase-multilingual-MiniLM-L12-v2
Maximum Sequence Length: 128 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': 128, '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})
)

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("philipp-zettl/MiniLM-similarity-small")
# Run inference
sentences = [
    'Change the currency for my payment',
    'payment query',
    'faq query',
]
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: MiniLM-dev
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.7357
spearman_cosine	0.7321
pearson_manhattan	0.624
spearman_manhattan	0.618
pearson_euclidean	0.6321
spearman_euclidean	0.6297
pearson_dot	0.7491
spearman_dot	0.7592
pearson_max	0.7491
spearman_max	0.7592

Semantic Similarity

Dataset: MiniLM-test
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.7687
spearman_cosine	0.7552
pearson_manhattan	0.7463
spearman_manhattan	0.7365
pearson_euclidean	0.7545
spearman_euclidean	0.7465
pearson_dot	0.7251
spearman_dot	0.739
pearson_max	0.7687
spearman_max	0.7552

Training Details

Training Dataset

Unnamed Dataset

Size: 844 training samples
Columns: sentence1, sentence2, and score
Approximate statistics based on the first 1000 samples:
sentence1 sentence2 score
type string string float
details
min: 6 tokens
mean: 10.8 tokens
max: 19 tokens

min: 4 tokens
mean: 5.33 tokens
max: 6 tokens

min: 0.0
mean: 0.49
max: 1.0

	sentence1	sentence2	score
type	string	string	float
details	min: 6 tokens mean: 10.8 tokens max: 19 tokens	min: 4 tokens mean: 5.33 tokens max: 6 tokens	min: 0.0 mean: 0.49 max: 1.0

Samples:

sentence1	sentence2	score
`Update the payment method for my order`	`order query`	`1.0`
`Не могу установить новое обновление, помогите!`	`support query`	`1.0`
`Помогите мне изменить настройки конфиденциальности`	`support query`	`1.0`

Loss: CoSENTLoss with these parameters:

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

Evaluation Dataset

Unnamed Dataset

Size: 106 evaluation samples
Columns: sentence1, sentence2, and score
Approximate statistics based on the first 1000 samples:
sentence1 sentence2 score
type string string float
details
min: 6 tokens
mean: 10.79 tokens
max: 15 tokens

min: 4 tokens
mean: 5.27 tokens
max: 6 tokens

min: 0.0
mean: 0.51
max: 1.0

	sentence1	sentence2	score
type	string	string	float
details	min: 6 tokens mean: 10.79 tokens max: 15 tokens	min: 4 tokens mean: 5.27 tokens max: 6 tokens	min: 0.0 mean: 0.51 max: 1.0

Samples:

sentence1	sentence2	score
`帮我修复系统错误`	`support query`	`1.0`
`Je veux commander une pizza`	`product query`	`1.0`
`Fix problems with my device’s Bluetooth connection`	`technical support query`	`1.0`

Loss: CoSENTLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
learning_rate: 2e-05
num_train_epochs: 2
warmup_ratio: 0.1
fp16: 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: 8
per_device_eval_batch_size: 8
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
eval_accumulation_steps: None
learning_rate: 2e-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: 0.1
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: 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: 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_eval_metrics: False
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss	MiniLM-dev_spearman_cosine	MiniLM-test_spearman_cosine
0.0943	10	4.0771	2.2054	0.2529	-
0.1887	20	4.4668	1.8221	0.3528	-
0.2830	30	2.5459	1.5545	0.4638	-
0.3774	40	2.1926	1.3145	0.5569	-
0.4717	50	0.9001	1.1653	0.6285	-
0.5660	60	1.4049	1.0734	0.6834	-
0.6604	70	0.7204	0.9951	0.6988	-
0.7547	80	1.4023	1.1213	0.6945	-
0.8491	90	0.2315	1.2931	0.6414	-
0.9434	100	0.0018	1.3904	0.6180	-
1.0377	110	0.0494	1.2889	0.6322	-
1.1321	120	0.3156	1.2461	0.6402	-
1.2264	130	1.8153	1.0844	0.6716	-
1.3208	140	0.2638	0.9939	0.6957	-
1.4151	150	0.5454	0.9545	0.7056	-
1.5094	160	0.3421	0.9699	0.7062	-
1.6038	170	0.0035	0.9521	0.7093	-
1.6981	180	0.0401	0.8988	0.7160	-
1.7925	190	0.8138	0.8619	0.7271	-
1.8868	200	0.0236	0.8449	0.7315	-
1.9811	210	0.0012	0.8438	0.7321	-
2.0	212	-	-	-	0.7552

Framework Versions

Python: 3.10.14
Sentence Transformers: 3.0.1
Transformers: 4.41.2
PyTorch: 2.3.1+cu121
Accelerate: 0.33.0
Datasets: 2.21.0
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",
}

CoSENTLoss

@online{kexuefm-8847,
    title={CoSENT: A more efficient sentence vector scheme than Sentence-BERT},
    author={Su Jianlin},
    year={2022},
    month={Jan},
    url={https://kexue.fm/archives/8847},
}