metadata

base_model: sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
datasets: []
language: []
library_name: sentence-transformers
metrics:
  - cosine_accuracy
  - dot_accuracy
  - manhattan_accuracy
  - euclidean_accuracy
  - max_accuracy
pipeline_tag: sentence-similarity
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:2400
  - loss:TripletLoss
  - loss:MultipleNegativesRankingLoss
  - loss:CoSENTLoss
widget:
  - source_sentence: Flislegging av hall
    sentences:
      - 'query: tapetsering av rom med grunnflate 4x4.5 meter minus tre dører'
      - 'query: fliser i hall'
      - 'query: fornye markiseduk'
  - source_sentence: Betongskjæring av rømningsvindu
    sentences:
      - Installere ventilasjonssystem
      - Installere nytt vindu i trevegg
      - Skjære ut rømningsvindu i betongvegg
  - source_sentence: Ny garasje leddport
    sentences:
      - Installere garasjeport
      - Bygge ny garasje
      - Legge nytt tak
  - source_sentence: Legge varmefolie i gang og stue.
    sentences:
      - Strø grusveier med salt
      - Legge varmekabler
      - Installere gulvvarme
  - source_sentence: Oppgradere kjeller til boareale
    sentences:
      - Oppussing av kjeller for boligformål
      - elektriker på bolig på 120kvm
      - Installere dusjkabinett
model-index:
  - name: >-
      SentenceTransformer based on
      sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
    results:
      - task:
          type: triplet
          name: Triplet
        dataset:
          name: test triplet evaluation
          type: test-triplet-evaluation
        metrics:
          - type: cosine_accuracy
            value: 0.7470049330514447
            name: Cosine Accuracy
          - type: dot_accuracy
            value: 0.31853417899929526
            name: Dot Accuracy
          - type: manhattan_accuracy
            value: 0.740662438336857
            name: Manhattan Accuracy
          - type: euclidean_accuracy
            value: 0.7420718816067653
            name: Euclidean Accuracy
          - type: max_accuracy
            value: 0.7470049330514447
            name: Max Accuracy

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("ostoveland/test3")
# Run inference
sentences = [
    'Oppgradere kjeller til boareale',
    'Oppussing av kjeller for boligformål',
    'Installere dusjkabinett',
]
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

Triplet

Dataset: test-triplet-evaluation
Evaluated with TripletEvaluator

Metric	Value
cosine_accuracy	0.747
dot_accuracy	0.3185
manhattan_accuracy	0.7407
euclidean_accuracy	0.7421
max_accuracy	0.747

Training Details

Training Datasets

Unnamed Dataset

Size: 800 training samples
Columns: sentence_0, sentence_1, and sentence_2

Approximate statistics based on the first 1000 samples:

	sentence_0	sentence_1	sentence_2
type	string	string	string
details	min: 3 tokens mean: 9.91 tokens max: 35 tokens	min: 4 tokens mean: 7.87 tokens max: 17 tokens	min: 4 tokens mean: 7.14 tokens max: 31 tokens

Samples:

sentence_0	sentence_1	sentence_2
`Oppussing av stue`	`Renovere stue`	`Male stue`
`Sameie søker vaktmestertjenester`	`Trenger vaktmester til sameie`	`Renholdstjenester for sameie`
`Sprenge og klargjøre til garasje`	`Grave ut til garasje`	`Bygge garasje`

Loss: TripletLoss with these parameters:

{
    "distance_metric": "TripletDistanceMetric.EUCLIDEAN",
    "triplet_margin": 5
}

Unnamed Dataset

Size: 800 training samples
Columns: sentence_0 and sentence_1
Approximate statistics based on the first 1000 samples:
sentence_0 sentence_1
type string string
details
min: 4 tokens
mean: 10.36 tokens
max: 35 tokens

min: 7 tokens
mean: 12.36 tokens
max: 26 tokens

	sentence_0	sentence_1
type	string	string
details	min: 4 tokens mean: 10.36 tokens max: 35 tokens	min: 7 tokens mean: 12.36 tokens max: 26 tokens

Samples:

sentence_0	sentence_1
`Helsparkle rom med totale veggflater på ca 20 m2`	`query: helsparkling av rom med 20 m2 veggflater`
`Reparere skifer tak og tak vindu`	`query: fikse takvindu og skifertak`
`Pigge opp flisgulv, fjerne gips vegger og gipstak - 11 kvm`	`query: fjerne flisgulv, gipsvegger og gipstak på 11 kvm`

Loss: MultipleNegativesRankingLoss with these parameters:

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

Unnamed Dataset

Size: 800 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: 4 tokens
mean: 10.32 tokens
max: 24 tokens

min: 4 tokens
mean: 8.18 tokens
max: 20 tokens

min: 0.1
mean: 0.51
max: 0.95

	sentence_0	sentence_1	label
type	string	string	float
details	min: 4 tokens mean: 10.32 tokens max: 24 tokens	min: 4 tokens mean: 8.18 tokens max: 20 tokens	min: 0.1 mean: 0.51 max: 0.95

Samples:

sentence_0	sentence_1	label
`Legging av våtromsbelegg`	`Renovering av bad`	`0.65`
`overvåkingskamera 3stk`	`installasjon av 3 overvåkingskameraer`	`0.95`
`Bytte lamper i portrom`	`Male portrom`	`0.15`

Loss: CoSENTLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 32
per_device_eval_batch_size: 32
num_train_epochs: 1
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: 32
per_device_eval_batch_size: 32
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: 1
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
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, '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: batch_sampler
multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	test-triplet-evaluation_max_accuracy
1.0	75	0.7470

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.0.1
Transformers: 4.41.2
PyTorch: 2.3.0+cu121
Accelerate: 0.31.0
Datasets: 2.20.0
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",
}

TripletLoss

@misc{hermans2017defense,
    title={In Defense of the Triplet Loss for Person Re-Identification}, 
    author={Alexander Hermans and Lucas Beyer and Bastian Leibe},
    year={2017},
    eprint={1703.07737},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

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}
}

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},
}