yahyaabd/indoSBERT-Large-mnrl-2

SentenceTransformer based on denaya/indoSBERT-large

This is a sentence-transformers model finetuned from denaya/indoSBERT-large on the statictable-triplets-all dataset. It maps sentences & paragraphs to a 256-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: denaya/indoSBERT-large
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 256 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • statictable-triplets-all

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': 1024, '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): Dense({'in_features': 1024, 'out_features': 256, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
)

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/indoSBERT-Large-mnrl-2")
# Run inference
sentences = [
    'Upah bulanan rata-rata: Hubungan pendidikan tertinggi dan sektor pekerjaan utama, data 2021',
    'Rata-rata Upah/Gaji Bersih sebulan Buruh/Karyawan Pegawai Menurut Pendidikan Tertinggi dan Lapangan Pekerjaan Utama, 2021',
    'Rata-rata Upah/Gaji Bersih Sebulan Buruh/Karyawan Pegawai Menurut Pendidikan Tertinggi dan Jenis Pekerjaan Utama, 2021',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 256]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Information Retrieval

• Dataset: bps-statictable-ir
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.9218
cosine_accuracy@5	0.9902
cosine_accuracy@10	0.9967
cosine_precision@1	0.9218
cosine_precision@5	0.2248
cosine_precision@10	0.1316
cosine_recall@1	0.7225
cosine_recall@5	0.793
cosine_recall@10	0.8182
cosine_ndcg@1	0.9218
cosine_ndcg@5	0.8341
cosine_ndcg@10	0.8332
cosine_mrr@1	0.9218
cosine_mrr@5	0.9522
cosine_mrr@10	0.9532
cosine_map@1	0.9218
cosine_map@5	0.792
cosine_map@10	0.7848

Training Details

Training Dataset

statictable-triplets-all

• Dataset: statictable-triplets-all at 24979b4
• Size: 967,831 training samples
• Columns: query, pos, and neg
• Approximate statistics based on the first 1000 samples:

  	query	pos	neg
  type	string	string	string
  details	min: 3 tokens mean: 16.97 tokens max: 30 tokens	min: 3 tokens mean: 20.79 tokens max: 48 tokens	min: 4 tokens mean: 20.94 tokens max: 48 tokens

• Samples:

  query	pos	neg
  Data input-output antar daerah, 34 provinsi: Transaksi domestik (52 industri, harga produsen, 2016)	Tabel Inter Regional Input-Output Indonesia Transaksi Domestik Atas Dasar Harga Produsen Menurut 34 Provinsi dan 52 Industri, 2016 (Juta Rupiah)	Penduduk Berumur 15 Tahun Ke Atas yang Bekerja Selama Seminggu yang Lalu Menurut Golongan Umur dan Jumlah Jam Kerja Seluruhnya, 2008 - 2024
  Data total penghasilan berbagai golongan rumah tangga setelah dipotong pajak, tahun 2000 (dalam miliar rupiah)	Jumlah Pendapatan Setelah Pajak Menurut Golongan Rumah Tangga (miliar rupiah), 2000, 2005, dan 2008	Institusi Korporasi Non Finansial Neraca Institusi Terintegrasi ( triliun rupiah), 2016 2022
  Rumah tangga dengan area resapan, data per provinsi, 2014	Persentase Rumah Tangga Menurut Provinsi dan Keberadaan Area Resapan Air, 2013-2014	Nilai Produksi dan Biaya Produksi per Musim Tanam per Hektar Budidaya Tanaman Padi Sawah, Padi Ladang, Jagung, dan Kedelai, 2017

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Evaluation Dataset

statictable-triplets-all

• Dataset: statictable-triplets-all at 24979b4
• Size: 967,831 evaluation samples
• Columns: query, pos, and neg
• Approximate statistics based on the first 1000 samples:

  	query	pos	neg
  type	string	string	string
  details	min: 3 tokens mean: 16.69 tokens max: 32 tokens	min: 3 tokens mean: 20.85 tokens max: 48 tokens	min: 3 tokens mean: 20.9 tokens max: 48 tokens

• Samples:

  query	pos	neg
  Kredit UMKM bank umum (miliar rupiah), 2012-2016	Posisi Kredit Usaha Mikro, Kecil, dan Menengah (UMKM) 1 pada Bank Umum (miliar rupiah), 2012-2016	Jumlah Penghuni Lapas per Kanwil
  Infant Mortality Rate di Indonesia per provinsi, 1971	Angka Kematian Bayi/AKB (Infant Mortality Rate/IMR) Menurut Provinsi, 1971-2020	Jumlah Sekolah, Guru, dan Murid Sekolah Menengah Kejuruan (SMK) di Bawah Kementrian Pendidikan dan Kebudayaan Menurut Provinsi tahun ajaran 2011/2012-2015/2016
  Partisipasi sekolah anak dan remaja: Data persentase usia 7-24 tahun per gender dan kelompok umur, 2021	Persentase Penduduk Usia 7-24 Tahun Menurut Jenis Kelamin, Kelompok Umur, dan Partisipasi Sekolah, 2002-2023	Tabel Input-Output Indonesia Transaksi Total Atas Dasar Harga Pembeli (17 Produk), 2016 (Juta Rupiah)

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 64
• per_device_eval_batch_size: 64
• num_train_epochs: 1
• warmup_ratio: 0.1
• fp16: True
• load_best_model_at_end: True
• eval_on_start: 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: 64
• per_device_eval_batch_size: 64
• 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: 1
• 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: 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: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	Validation Loss	bps-statictable-ir_cosine_ndcg@10
0	0	-	0.7678	0.7378
0.1391	100	0.2164	0.0292	0.8324
0.2782	200	0.032	0.0143	0.8383
0.4172	300	0.0221	0.0077	0.8392
0.5563	400	0.0088	0.0055	0.8391
0.6954	500	0.0058	0.0033	0.8301
0.8345	600	0.0039	0.0016	0.8331
0.9736	700	0.0027	0.0019	0.8332

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.11.11
• Sentence Transformers: 3.4.1
• Transformers: 4.48.3
• PyTorch: 2.6.0+cu124
• Accelerate: 1.3.0
• Datasets: 3.4.1
• Tokenizers: 0.21.1

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",
}

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}
}