yahyaabd/allstats-search-miniLM-v1-5

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:
  • query-hard-pos-neg-doc-pairs-statictable

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-v1-5")
# 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 and allstats-semantic-mini-v1_dev
• Evaluated with BinaryClassificationEvaluator

Metric	allstats-semantic-mini-v1_test	allstats-semantic-mini-v1_dev
cosine_accuracy	0.977	0.977
cosine_accuracy_threshold	0.747	0.747
cosine_f1	0.9649	0.9649
cosine_f1_threshold	0.7452	0.7452
cosine_precision	0.9553	0.9553
cosine_recall	0.9746	0.9746
cosine_ap	0.9927	0.9927
cosine_mcc	0.9479	0.9479

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, and label
• 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, and label
• 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: steps
• per_device_train_batch_size: 24
• per_device_eval_batch_size: 24
• num_train_epochs: 2
• warmup_ratio: 0.2
• fp16: True
• load_best_model_at_end: True
• eval_on_start: True

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 24
• per_device_eval_batch_size: 24
• 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: 1.0
• num_train_epochs: 2
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.2
• 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: True
• 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: None
• hub_always_push: False
• 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
• 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
• eval_on_start: True
• use_liger_kernel: False
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_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	-	-	0.8789	-
0	0	-	0.7267	-	0.8789
0.0188	20	0.668	0.6453	-	0.8848
0.0375	40	0.6117	0.4411	-	0.9003
0.0563	60	0.3108	0.3592	-	0.9130
0.0750	80	0.3824	0.2899	-	0.9336
0.0938	100	0.2118	0.2530	-	0.9442
0.1126	120	0.232	0.1945	-	0.9582
0.1313	140	0.1233	0.1663	-	0.9656
0.1501	160	0.1293	0.1655	-	0.9654
0.1689	180	0.0714	0.2142	-	0.9578
0.1876	200	0.1198	0.1455	-	0.9702
0.2064	220	0.1081	0.1258	-	0.9766
0.2251	240	0.0484	0.1210	-	0.9753
0.2439	260	0.1463	0.1100	-	0.9792
0.2627	280	0.0422	0.1228	-	0.9777
0.2814	300	0.1187	0.1302	-	0.9725
0.3002	320	0.0635	0.1257	-	0.9733
0.3189	340	0.0422	0.1125	-	0.9736
0.3377	360	0.0479	0.0882	-	0.9796
0.3565	380	0.119	0.1319	-	0.9697
0.3752	400	0.099	0.1445	-	0.9702
0.3940	420	0.0409	0.1434	-	0.9706
0.4128	440	0.1053	0.1520	-	0.9686
0.4315	460	0.1035	0.1382	-	0.9727
0.4503	480	0.0848	0.1150	-	0.9789
0.4690	500	0.0387	0.0944	-	0.9826
0.4878	520	0.0097	0.1041	-	0.9811
0.5066	540	0.0667	0.1041	-	0.9783
0.5253	560	0.1028	0.1386	-	0.9736
0.5441	580	0.0543	0.1350	-	0.9769
0.5629	600	0.0859	0.1254	-	0.9776
0.5816	620	0.0853	0.1483	-	0.9728
0.6004	640	0.024	0.1159	-	0.9781
0.6191	660	0.0762	0.1046	-	0.9784
0.6379	680	0.0433	0.1275	-	0.9686
0.6567	700	0.0772	0.0592	-	0.9882
0.6754	720	0.0185	0.0542	-	0.9889
0.6942	740	0.0376	0.1123	-	0.9801
0.7129	760	0.0612	0.1002	-	0.9817
0.7317	780	0.0156	0.0948	-	0.9809
0.7505	800	0.0474	0.0778	-	0.9817
0.7692	820	0.0427	0.0824	-	0.9828
0.7880	840	0.0289	0.0911	-	0.9833
0.8068	860	0.0175	0.0991	-	0.9827
0.8255	880	0.0241	0.0951	-	0.9824
0.8443	900	0.0527	0.0816	-	0.9860
0.8630	920	0.0535	0.0707	-	0.9875
0.8818	940	0.0211	0.0767	-	0.9868
0.9006	960	0.013	0.0758	-	0.9872
0.9193	980	0.0079	0.0781	-	0.9848
0.9381	1000	0.0406	0.0820	-	0.9845
0.9568	1020	0.0277	0.0685	-	0.9874
0.9756	1040	0.0132	0.0760	-	0.9859
0.9944	1060	0.0268	0.0881	-	0.9833
1.0131	1080	0.0089	0.0772	-	0.9857
1.0319	1100	0.0276	0.0773	-	0.9850
1.0507	1120	0.0181	0.0729	-	0.9860
1.0694	1140	0.0065	0.0683	-	0.9867
1.0882	1160	0.01	0.0639	-	0.9873
1.1069	1180	0.0068	0.0662	-	0.9870
1.1257	1200	0.0	0.0722	-	0.9863
1.1445	1220	0.0067	0.0710	-	0.9866
1.1632	1240	0.0069	0.0666	-	0.9877
1.1820	1260	0.0	0.0639	-	0.9880
1.2008	1280	0.0244	0.0610	-	0.9882
1.2195	1300	0.0143	0.0630	-	0.9877
1.2383	1320	0.0173	0.0530	-	0.9896
1.2570	1340	0.0171	0.0496	-	0.9907
1.2758	1360	0.0225	0.0521	-	0.9909
1.2946	1380	0.011	0.0569	-	0.9900
1.3133	1400	0.0088	0.0605	-	0.9898
1.3321	1420	0.0	0.0619	-	0.9897
1.3508	1440	0.0135	0.0608	-	0.9894
1.3696	1460	0.0	0.0593	-	0.9892
1.3884	1480	0.0145	0.0578	-	0.9894
1.4071	1500	0.0	0.0608	-	0.9896
1.4259	1520	0.0069	0.0567	-	0.9906
1.4447	1540	0.0	0.0561	-	0.9907
1.4634	1560	0.0224	0.0531	-	0.9912
1.4822	1580	0.0	0.0523	-	0.9911
1.5009	1600	0.0066	0.0503	-	0.9912
1.5197	1620	0.0	0.0472	-	0.9915
1.5385	1640	0.018	0.0452	-	0.9923
1.5572	1660	0.0117	0.0449	-	0.9925
1.5760	1680	0.0	0.0456	-	0.9925
1.5947	1700	0.0	0.0448	-	0.9925
1.6135	1720	0.0	0.0448	-	0.9925
1.6323	1740	0.0072	0.0458	-	0.9924
1.6510	1760	0.0	0.0456	-	0.9923
1.6698	1780	0.0163	0.0482	-	0.9925
1.6886	1800	0.0063	0.0463	-	0.9926
1.7073	1820	0.0078	0.0482	-	0.9925
1.7261	1840	0.0179	0.0472	-	0.9927
1.7448	1860	0.0	0.0477	-	0.9927
1.7636	1880	0.0	0.0477	-	0.9927
1.7824	1900	0.0065	0.0461	-	0.9926
1.8011	1920	0.0077	0.0458	-	0.9926
1.8199	1940	0.0065	0.0453	-	0.9927
1.8386	1960	0.0	0.0451	-	0.9927
1.8574	1980	0.0	0.0451	-	0.9927
1.8762	2000	0.0	0.0451	-	0.9927
1.8949	2020	0.0	0.0451	-	0.9927
1.9137	2040	0.0	0.0451	-	0.9927
1.9325	2060	0.0	0.0451	-	0.9927
1.9512	2080	0.0	0.0451	-	0.9927
1.9700	2100	0.007	0.0442	-	0.9927
1.9887	2120	0.0067	0.0441	-	0.9927
-1	-1	-	-	0.9927	-

• 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",
}