x2bee/KoModernBERT-base-nli-sts-SBERT_v01

SentenceTransformer based on x2bee/ModernBert_MLM_kotoken_v03

This is a sentence-transformers model finetuned from x2bee/ModernBert_MLM_kotoken_v03 on the misc_sts_pairs_v2_kor dataset. It maps sentences & paragraphs to a 768-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: x2bee/ModernBert_MLM_kotoken_v03
• Maximum Sequence Length: 8192 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • misc_sts_pairs_v2_kor

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': 8192, 'do_lower_case': False}) with Transformer model: ModernBertModel 
  (1): Pooling({'word_embedding_dimension': 768, '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("x2bee/KoModernBERT-base-nli-sts-SBERT_v01")
# Run inference
sentences = [
    '수동 운전석 창문을 어떻게 수리하나요?',
    '1992년형 혼다 시빅에서 올라가지 않는 수동 창문을 어떻게 수리하나요?',
    '아홉 번째 닥터가 멈춘 닥터 후 에피소드는 무엇입니까?',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

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

Evaluation

Metrics

Semantic Similarity

• Dataset: sts_dev
• Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.524
spearman_cosine	0.5139
pearson_euclidean	0.5051
spearman_euclidean	0.5001
pearson_manhattan	0.5087
spearman_manhattan	0.504
pearson_dot	0.4545
spearman_dot	0.4439
pearson_max	0.524
spearman_max	0.5139

Training Details

Training Dataset

misc_sts_pairs_v2_kor

• Dataset: misc_sts_pairs_v2_kor at 845f810
• Size: 449,904 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: 17.81 tokens max: 49 tokens	min: 6 tokens mean: 17.78 tokens max: 80 tokens	min: 0.53 mean: 0.75 max: 0.98

• Samples:

  sentence1	sentence2	score
  1999년형 유콘 4륜구동 차량의 앞쪽 조수석 타이어에서 발생하는 갈리는 소음의 원인은 무엇인가요?	차의 오른쪽 앞쪽에서 발생하는 갈리는 소리의 원인은 무엇인가요?	0.8193586337477191
  왜 제임스타운 정착민들은 그곳의 원주민들과 갈등을 겪었는가?	왜 제임스타운은 원주민들과 갈등을 겪었는가?	0.8701910827908218
  옥수수 전분을 섭취하는 것이 건강에 어떤 영향을 미칠 수 있습니까?	옥수수 전분을 섭취하면 당신에게 어떤 영향을 미칠까요?	0.8809354609563622

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Evaluation Dataset

misc_sts_pairs_v2_kor

• Dataset: misc_sts_pairs_v2_kor at 845f810
• Size: 449,904 evaluation samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 7 tokens mean: 17.76 tokens max: 65 tokens	min: 6 tokens mean: 17.65 tokens max: 52 tokens	min: 0.53 mean: 0.75 max: 0.98

• Samples:

  sentence1	sentence2	score
  용광로의 온도는 얼마나 뜨거운가?	용광로의 온도는 얼마나 높습니까?	0.751853250408994
  영어로 'Lei è il mio uno e solo'는 어떻게 철자하나요?	'Lei è il mio uno e solo'의 영어 동등어는 무엇인가요?	0.8265661603331053
  버드와이저 포커 광고에 나오는 소녀는 누구인가요?	포커 스타일의 버드와이저 광고에 나오는 소녀는 누구인가요?	0.9301912848973812

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• gradient_accumulation_steps: 4
• learning_rate: 1e-05
• num_train_epochs: 2
• warmup_ratio: 0.3
• push_to_hub: True
• hub_model_id: x2bee/KoModernBERT-base-nli-sts-SBERT_v01
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: epoch
• 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: 4
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 1e-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: 0.3
• 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: True
• 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: True
• resume_from_checkpoint: None
• hub_model_id: x2bee/KoModernBERT-base-nli-sts-SBERT_v01
• 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: False
• 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	sts_dev_spearman_max
0	0	-	-	0.5070
0.2397	100	0.0311	-	-
0.4793	200	0.0082	-	-
0.7190	300	0.0065	-	-
0.9587	400	0.0061	-	-
1.0	418	-	0.0059	0.4899
1.1965	500	0.0058	-	-
1.4362	600	0.0057	-	-
1.6759	700	0.0055	-	-
1.9155	800	0.0053	-	-
1.9970	834	-	0.0057	0.5139

Framework Versions

• Python: 3.11.10
• Sentence Transformers: 3.3.1
• Transformers: 4.48.0.dev0
• PyTorch: 2.5.1+cu124
• Accelerate: 1.2.1
• 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",
}