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 on the query-hard-pos-neg-doc-pairs-statictable dataset. 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 dimensions
- Similarity Function: Cosine Similarity
- Training Dataset:
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("yahyaabd/allstats-search-miniLM-v1-6")
# Run inference
sentences = [
'Arus dana Q3 2006',
'Ringkasan Neraca Arus Dana, Triwulan III, 2006, (Miliar Rupiah)',
'Rata-Rata Pengeluaran per Kapita Sebulan di Daerah Perkotaan Menurut Kelompok Barang dan Golongan Pengeluaran per Kapita Sebulan, 2000-2012',
]
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
Binary Classification
- Datasets:
allstats-semantic-mini-v1_test
andallstats-semantic-mini-v1_dev
- Evaluated with
BinaryClassificationEvaluator
Metric | allstats-semantic-mini-v1_test | allstats-semantic-mini-v1_dev |
---|---|---|
cosine_accuracy | 0.9809 | 0.9754 |
cosine_accuracy_threshold | 0.7711 | 0.7737 |
cosine_f1 | 0.9706 | 0.9623 |
cosine_f1_threshold | 0.7711 | 0.7737 |
cosine_precision | 0.9725 | 0.9545 |
cosine_recall | 0.9687 | 0.9701 |
cosine_ap | 0.9957 | 0.9927 |
cosine_mcc | 0.9565 | 0.9441 |
Training Details
Training Dataset
query-hard-pos-neg-doc-pairs-statictable
- Dataset: query-hard-pos-neg-doc-pairs-statictable at 7b28b96
- Size: 25,580 training samples
- Columns:
query
,doc
, andlabel
- Approximate statistics based on the first 1000 samples:
query doc label type string string int details - min: 7 tokens
- mean: 20.14 tokens
- max: 55 tokens
- min: 5 tokens
- mean: 24.9 tokens
- max: 47 tokens
- 0: ~70.80%
- 1: ~29.20%
- Samples:
query doc label Status pekerjaan utama penduduk usia 15+ yang bekerja, 2020
Jumlah Penghuni Lapas per Kanwil
0
status pekerjaan utama penduduk usia 15+ yang bekerja, 2020
Jumlah Penghuni Lapas per Kanwil
0
STATUS PEKERJAAN UTAMA PENDUDUK USIA 15+ YANG BEKERJA, 2020
Jumlah Penghuni Lapas per Kanwil
0
- Loss:
OnlineContrastiveLoss
Evaluation Dataset
query-hard-pos-neg-doc-pairs-statictable
- Dataset: query-hard-pos-neg-doc-pairs-statictable at 7b28b96
- Size: 5,479 evaluation samples
- Columns:
query
,doc
, andlabel
- Approximate statistics based on the first 1000 samples:
query doc label type string string int details - min: 7 tokens
- mean: 20.78 tokens
- max: 52 tokens
- min: 4 tokens
- mean: 26.28 tokens
- max: 43 tokens
- 0: ~71.50%
- 1: ~28.50%
- Samples:
query doc label Bagaimana perbandingan PNS pria dan wanita di berbagai golongan tahun 2014?
Rata-rata Pendapatan Bersih Berusaha Sendiri Menurut Provinsi dan Lapangan Pekerjaan Utama (ribu rupiah), 2017
0
bagaimana perbandingan pns pria dan wanita di berbagai golongan tahun 2014?
Rata-rata Pendapatan Bersih Berusaha Sendiri Menurut Provinsi dan Lapangan Pekerjaan Utama (ribu rupiah), 2017
0
BAGAIMANA PERBANDINGAN PNS PRIA DAN WANITA DI BERBAGAI GOLONGAN TAHUN 2014?
Rata-rata Pendapatan Bersih Berusaha Sendiri Menurut Provinsi dan Lapangan Pekerjaan Utama (ribu rupiah), 2017
0
- Loss:
OnlineContrastiveLoss
Training Hyperparameters
Non-Default Hyperparameters
eval_strategy
: stepsper_device_train_batch_size
: 32per_device_eval_batch_size
: 32warmup_ratio
: 0.2fp16
: Trueload_best_model_at_end
: Trueeval_on_start
: True
All Hyperparameters
Click to expand
overwrite_output_dir
: Falsedo_predict
: Falseeval_strategy
: stepsprediction_loss_only
: Trueper_device_train_batch_size
: 32per_device_eval_batch_size
: 32per_gpu_train_batch_size
: Noneper_gpu_eval_batch_size
: Nonegradient_accumulation_steps
: 1eval_accumulation_steps
: Nonetorch_empty_cache_steps
: Nonelearning_rate
: 5e-05weight_decay
: 0.0adam_beta1
: 0.9adam_beta2
: 0.999adam_epsilon
: 1e-08max_grad_norm
: 1.0num_train_epochs
: 3max_steps
: -1lr_scheduler_type
: linearlr_scheduler_kwargs
: {}warmup_ratio
: 0.2warmup_steps
: 0log_level
: passivelog_level_replica
: warninglog_on_each_node
: Truelogging_nan_inf_filter
: Truesave_safetensors
: Truesave_on_each_node
: Falsesave_only_model
: Falserestore_callback_states_from_checkpoint
: Falseno_cuda
: Falseuse_cpu
: Falseuse_mps_device
: Falseseed
: 42data_seed
: Nonejit_mode_eval
: Falseuse_ipex
: Falsebf16
: Falsefp16
: Truefp16_opt_level
: O1half_precision_backend
: autobf16_full_eval
: Falsefp16_full_eval
: Falsetf32
: Nonelocal_rank
: 0ddp_backend
: Nonetpu_num_cores
: Nonetpu_metrics_debug
: Falsedebug
: []dataloader_drop_last
: Falsedataloader_num_workers
: 0dataloader_prefetch_factor
: Nonepast_index
: -1disable_tqdm
: Falseremove_unused_columns
: Truelabel_names
: Noneload_best_model_at_end
: Trueignore_data_skip
: Falsefsdp
: []fsdp_min_num_params
: 0fsdp_config
: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}fsdp_transformer_layer_cls_to_wrap
: Noneaccelerator_config
: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}deepspeed
: Nonelabel_smoothing_factor
: 0.0optim
: adamw_torchoptim_args
: Noneadafactor
: Falsegroup_by_length
: Falselength_column_name
: lengthddp_find_unused_parameters
: Noneddp_bucket_cap_mb
: Noneddp_broadcast_buffers
: Falsedataloader_pin_memory
: Truedataloader_persistent_workers
: Falseskip_memory_metrics
: Trueuse_legacy_prediction_loop
: Falsepush_to_hub
: Falseresume_from_checkpoint
: Nonehub_model_id
: Nonehub_strategy
: every_savehub_private_repo
: Nonehub_always_push
: Falsegradient_checkpointing
: Falsegradient_checkpointing_kwargs
: Noneinclude_inputs_for_metrics
: Falseinclude_for_metrics
: []eval_do_concat_batches
: Truefp16_backend
: autopush_to_hub_model_id
: Nonepush_to_hub_organization
: Nonemp_parameters
:auto_find_batch_size
: Falsefull_determinism
: Falsetorchdynamo
: Noneray_scope
: lastddp_timeout
: 1800torch_compile
: Falsetorch_compile_backend
: Nonetorch_compile_mode
: Nonedispatch_batches
: Nonesplit_batches
: Noneinclude_tokens_per_second
: Falseinclude_num_input_tokens_seen
: Falseneftune_noise_alpha
: Noneoptim_target_modules
: Nonebatch_eval_metrics
: Falseeval_on_start
: Trueuse_liger_kernel
: Falseeval_use_gather_object
: Falseaverage_tokens_across_devices
: Falseprompts
: Nonebatch_sampler
: batch_samplermulti_dataset_batch_sampler
: proportional
Training Logs
Click to expand
Epoch | Step | Training Loss | Validation Loss | allstats-semantic-mini-v1_test_cosine_ap | allstats-semantic-mini-v1_dev_cosine_ap |
---|---|---|---|---|---|
-1 | -1 | - | - | 0.8910 | - |
0 | 0 | - | 1.0484 | - | 0.8789 |
0.025 | 20 | 1.0003 | 0.9175 | - | 0.8856 |
0.05 | 40 | 0.6667 | 0.6433 | - | 0.9010 |
0.075 | 60 | 0.5982 | 0.5203 | - | 0.9145 |
0.1 | 80 | 0.4476 | 0.4175 | - | 0.9344 |
0.125 | 100 | 0.3489 | 0.3152 | - | 0.9540 |
0.15 | 120 | 0.1643 | 0.2726 | - | 0.9602 |
0.175 | 140 | 0.2126 | 0.2525 | - | 0.9631 |
0.2 | 160 | 0.1797 | 0.2151 | - | 0.9715 |
0.225 | 180 | 0.1304 | 0.1895 | - | 0.9756 |
0.25 | 200 | 0.1714 | 0.2142 | - | 0.9767 |
0.275 | 220 | 0.1758 | 0.1840 | - | 0.9791 |
0.3 | 240 | 0.0562 | 0.1723 | - | 0.9801 |
0.325 | 260 | 0.0863 | 0.1656 | - | 0.9773 |
0.35 | 280 | 0.12 | 0.1806 | - | 0.9788 |
0.375 | 300 | 0.0982 | 0.1792 | - | 0.9769 |
0.4 | 320 | 0.0421 | 0.1724 | - | 0.9783 |
0.425 | 340 | 0.1078 | 0.2158 | - | 0.9733 |
0.45 | 360 | 0.0882 | 0.1501 | - | 0.9822 |
0.475 | 380 | 0.0251 | 0.1334 | - | 0.9843 |
0.5 | 400 | 0.0267 | 0.1238 | - | 0.9855 |
0.525 | 420 | 0.0899 | 0.1404 | - | 0.9859 |
0.55 | 440 | 0.0782 | 0.1253 | - | 0.9852 |
0.575 | 460 | 0.1209 | 0.1772 | - | 0.9768 |
0.6 | 480 | 0.0643 | 0.1817 | - | 0.9763 |
0.625 | 500 | 0.1051 | 0.2030 | - | 0.9748 |
0.65 | 520 | 0.0494 | 0.1405 | - | 0.9814 |
0.675 | 540 | 0.0548 | 0.1175 | - | 0.9831 |
0.7 | 560 | 0.121 | 0.1597 | - | 0.9819 |
0.725 | 580 | 0.0642 | 0.1675 | - | 0.9811 |
0.75 | 600 | 0.0618 | 0.1539 | - | 0.9827 |
0.775 | 620 | 0.0745 | 0.1149 | - | 0.9845 |
0.8 | 640 | 0.0452 | 0.1562 | - | 0.9797 |
0.825 | 660 | 0.0816 | 0.1580 | - | 0.9816 |
0.85 | 680 | 0.0957 | 0.1192 | - | 0.9830 |
0.875 | 700 | 0.06 | 0.1100 | - | 0.9863 |
0.9 | 720 | 0.018 | 0.1300 | - | 0.9822 |
0.925 | 740 | 0.0213 | 0.1267 | - | 0.9843 |
0.95 | 760 | 0.0263 | 0.1687 | - | 0.9796 |
0.975 | 780 | 0.032 | 0.1250 | - | 0.9849 |
1.0 | 800 | 0.065 | 0.1363 | - | 0.9828 |
1.025 | 820 | 0.0174 | 0.1394 | - | 0.9835 |
1.05 | 840 | 0.0568 | 0.1124 | - | 0.9849 |
1.075 | 860 | 0.0464 | 0.1174 | - | 0.9826 |
1.1 | 880 | 0.013 | 0.1178 | - | 0.9814 |
1.125 | 900 | 0.0331 | 0.1239 | - | 0.9812 |
1.15 | 920 | 0.0416 | 0.1240 | - | 0.9817 |
1.175 | 940 | 0.0111 | 0.1303 | - | 0.9840 |
1.2 | 960 | 0.0441 | 0.1156 | - | 0.9854 |
1.225 | 980 | 0.0243 | 0.0972 | - | 0.9879 |
1.25 | 1000 | 0.0 | 0.0917 | - | 0.9877 |
1.275 | 1020 | 0.0477 | 0.0863 | - | 0.9885 |
1.3 | 1040 | 0.0108 | 0.1029 | - | 0.9877 |
1.325 | 1060 | 0.0 | 0.1103 | - | 0.9869 |
1.35 | 1080 | 0.0134 | 0.1113 | - | 0.9871 |
1.375 | 1100 | 0.0 | 0.1146 | - | 0.9870 |
1.4 | 1120 | 0.0132 | 0.1218 | - | 0.9862 |
1.425 | 1140 | 0.0223 | 0.0948 | - | 0.9883 |
1.45 | 1160 | 0.0183 | 0.0883 | - | 0.9883 |
1.475 | 1180 | 0.0378 | 0.0961 | - | 0.9881 |
1.5 | 1200 | 0.0114 | 0.0961 | - | 0.9882 |
1.525 | 1220 | 0.0143 | 0.1020 | - | 0.9861 |
1.55 | 1240 | 0.0183 | 0.0867 | - | 0.9888 |
1.575 | 1260 | 0.0 | 0.0858 | - | 0.9892 |
1.6 | 1280 | 0.0 | 0.0858 | - | 0.9892 |
1.625 | 1300 | 0.0 | 0.0858 | - | 0.9892 |
1.65 | 1320 | 0.0172 | 0.0846 | - | 0.9896 |
1.675 | 1340 | 0.0153 | 0.0754 | - | 0.9917 |
1.7 | 1360 | 0.0163 | 0.0770 | - | 0.9913 |
1.725 | 1380 | 0.0167 | 0.0943 | - | 0.9901 |
1.75 | 1400 | 0.0148 | 0.0964 | - | 0.9899 |
1.775 | 1420 | 0.0065 | 0.0930 | - | 0.9902 |
1.8 | 1440 | 0.0 | 0.0945 | - | 0.9904 |
1.825 | 1460 | 0.0067 | 0.0991 | - | 0.9895 |
1.85 | 1480 | 0.0194 | 0.0996 | - | 0.9894 |
1.875 | 1500 | 0.0 | 0.0953 | - | 0.9903 |
1.9 | 1520 | 0.0236 | 0.0883 | - | 0.9906 |
1.925 | 1540 | 0.0111 | 0.0858 | - | 0.9904 |
1.95 | 1560 | 0.0 | 0.0878 | - | 0.9903 |
1.975 | 1580 | 0.0147 | 0.0849 | - | 0.9906 |
2.0 | 1600 | 0.0154 | 0.0852 | - | 0.9902 |
2.025 | 1620 | 0.0067 | 0.0861 | - | 0.9903 |
2.05 | 1640 | 0.019 | 0.0859 | - | 0.9907 |
2.075 | 1660 | 0.0083 | 0.0875 | - | 0.9908 |
2.1 | 1680 | 0.0067 | 0.0771 | - | 0.9917 |
2.125 | 1700 | 0.0 | 0.0773 | - | 0.9917 |
2.15 | 1720 | 0.0071 | 0.0771 | - | 0.9919 |
2.175 | 1740 | 0.0064 | 0.0756 | - | 0.9916 |
2.2 | 1760 | 0.0 | 0.0772 | - | 0.9916 |
2.225 | 1780 | 0.0 | 0.0772 | - | 0.9915 |
2.25 | 1800 | 0.0158 | 0.0734 | - | 0.9920 |
2.275 | 1820 | 0.0 | 0.0730 | - | 0.9920 |
2.3 | 1840 | 0.0 | 0.0733 | - | 0.9920 |
2.325 | 1860 | 0.0161 | 0.0681 | - | 0.9922 |
2.35 | 1880 | 0.0 | 0.0713 | - | 0.9920 |
2.375 | 1900 | 0.0 | 0.0721 | - | 0.9920 |
2.4 | 1920 | 0.0 | 0.0722 | - | 0.9920 |
2.425 | 1940 | 0.0064 | 0.0648 | - | 0.9928 |
2.45 | 1960 | 0.0068 | 0.0641 | - | 0.9930 |
2.475 | 1980 | 0.0069 | 0.0635 | - | 0.9929 |
2.5 | 2000 | 0.0066 | 0.0657 | - | 0.9929 |
2.525 | 2020 | 0.0 | 0.0657 | - | 0.9930 |
2.55 | 2040 | 0.0139 | 0.0657 | - | 0.9931 |
2.575 | 2060 | 0.0 | 0.0667 | - | 0.9931 |
2.6 | 2080 | 0.0 | 0.0666 | - | 0.9931 |
2.625 | 2100 | 0.0 | 0.0666 | - | 0.9931 |
2.65 | 2120 | 0.0 | 0.0666 | - | 0.9931 |
2.675 | 2140 | 0.0 | 0.0667 | - | 0.9931 |
2.7 | 2160 | 0.0 | 0.0666 | - | 0.9931 |
2.725 | 2180 | 0.0 | 0.0666 | - | 0.9931 |
2.75 | 2200 | 0.0071 | 0.0665 | - | 0.9931 |
2.775 | 2220 | 0.0 | 0.0671 | - | 0.9931 |
2.8 | 2240 | 0.0071 | 0.0692 | - | 0.9928 |
2.825 | 2260 | 0.0 | 0.0700 | - | 0.9927 |
2.85 | 2280 | 0.0068 | 0.0688 | - | 0.9927 |
2.875 | 2300 | 0.0 | 0.0688 | - | 0.9927 |
2.9 | 2320 | 0.0 | 0.0688 | - | 0.9927 |
2.925 | 2340 | 0.0 | 0.0688 | - | 0.9927 |
2.95 | 2360 | 0.0 | 0.0688 | - | 0.9927 |
2.975 | 2380 | 0.0 | 0.0688 | - | 0.9927 |
3.0 | 2400 | 0.0 | 0.0688 | - | 0.9927 |
-1 | -1 | - | - | 0.9957 | - |
- The bold row denotes the saved checkpoint.
Framework Versions
- Python: 3.10.12
- Sentence Transformers: 3.4.0
- Transformers: 4.48.1
- PyTorch: 2.5.1+cu124
- Accelerate: 1.3.0
- Datasets: 3.2.0
- Tokenizers: 0.21.0
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",
}
- Downloads last month
- 8
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
Model tree for yahyaabd/allstats-search-miniLM-v1-6
Dataset used to train yahyaabd/allstats-search-miniLM-v1-6
Evaluation results
- Cosine Accuracy on allstats semantic mini v1 testself-reported0.981
- Cosine Accuracy Threshold on allstats semantic mini v1 testself-reported0.771
- Cosine F1 on allstats semantic mini v1 testself-reported0.971
- Cosine F1 Threshold on allstats semantic mini v1 testself-reported0.771
- Cosine Precision on allstats semantic mini v1 testself-reported0.973
- Cosine Recall on allstats semantic mini v1 testself-reported0.969
- Cosine Ap on allstats semantic mini v1 testself-reported0.996
- Cosine Mcc on allstats semantic mini v1 testself-reported0.956
- Cosine Accuracy on allstats semantic mini v1 devself-reported0.975
- Cosine Accuracy Threshold on allstats semantic mini v1 devself-reported0.774