SentenceTransformer based on google/embeddinggemma-300m

This is a sentence-transformers model finetuned from google/embeddinggemma-300m. 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: google/embeddinggemma-300m
• Maximum Sequence Length: 2048 tokens
• Output Dimensionality: 768 dimensions
• 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': 2048, 'do_lower_case': False, 'architecture': 'Gemma3TextModel'})
  (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})
  (2): Dense({'in_features': 768, 'out_features': 3072, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
  (3): Dense({'in_features': 3072, 'out_features': 768, 'bias': False, 'activation_function': 'torch.nn.modules.linear.Identity'})
  (4): Normalize()
)

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("Day1Kim/embeddinggemma-300m-Korean")
# Run inference
queries = [
    "\uc5b5\uc9c0\ub9d0\ub69d\uc73c\ub85c \ubcf4\uac15\ub41c \ube44\ud0c8\uba74\uc758 \ub0b4\uc9c4\uc124\uacc4 \uacb0\uc815 \uae30\uc900\uc740 \ubb34\uc5c7\uc778\uac00\uc694?",
]
documents = [
    '비탈면보강공법 KDS117015:2020KDS110000지반설계기준 9입시켜야한다 .4.3.3내진설계여부(1)억지말뚝으로보강된비탈면의내진설계는보강되지않은비탈면의내진설계여부에따라결정하며 ,KDS 119000의비탈면내진등급을참고한다 .(2)억지말뚝으로보강된비탈면의지진시안정해석은 4.4및KDS 119000을참조한다 .4.4지진시안정해석4.4.1네일(1)지진시네일로보강된비탈면의안정해석에서는내적안정과외적안정성을검토한다 .(2)네일로보강된비탈면의지진시안정해석에서고려하는지진하중은파괴토체의자중과지진계수 (Am)를곱한등가지진력으로 고려하며 ,파괴토체의중심에횡방향으로작용시킨다 .(3)지진에의한지진계수는 KDS 119000(1.6.5)에서제시하는유효수평지반가 속도(S)를이용하여산정한다 .4.4.2록볼트(1)지진시록볼트로보강된비탈면의안정해석에서는외적안정성을검토한다 .(2)록볼트로보강된비탈면의지진시안정해석에서고려하는지진하중은파괴토체의자중과지진계수 (Am)를곱한등가지진력으로 고려하며 ,파괴토체의중심에횡방향으로작용시킨다 .',
    '지반계측 KDS111015:2021KDS110000지반설계기준 344.2.2.2 계측기기운용기법(1) 인력에의한계측기기운용과자동화장비에의한운용기법으로크게구분할수있으며, 붕괴및활동의진행특성, 계측대상시설물의중요도 , 피해발생시영향, 경제성, 계측빈도등을고려하여운용기법을선택하여야한다.4.2.2.3 일반적인계측관리의자동화(1) 기록지또는저장장치에계측자료를기록할때까지를자동화하고 , 그후의처리는별도로컴퓨터로실시하는반자동계측관리기법과 , 자료수집ㆍ해석ㆍ그래프화까지를유선ㆍ무선으로온라인화된시스템으로일관하여실시하는전자동계측관리기법및상기의두가지방법을병용하는기법으로구분하며 , 계측대상조건을고려하여운용기법을선정하여야한다.4.2.2.4 피해방지및최소화방법(1) 조기에징후를감지하는것이중요하고 , 모니터링과동시에신속하게그정보를전달ㆍ처리하는것이필요하며계측자료의수집ㆍ처리ㆍ해석까지를일괄하여처리하는자동화기술을사용하는것을고려하여야한다.',
    '(9)매립지반에설치하는기초는특별히주의하여야하는데 ,기초하부의접지압력은가급적균등하게분포하도록설계하고 ,접지압력의분포는흙또는암,그리고구조물의특성에따라달라지며 ,토질역학및암석역학적원리들과일치하여야한다 .2.조사및계획∙내용없음 .3.재료∙내용없음 .4.설계4.1지지력산정4.1.1지지력산정을위한고려사항(1)기초설계시시추조사 ,현장및실내시험을통하여지반특성을파악한후지지력을산정한다 .그러나상재하중이작은구조물또는가설구조물의기초는인근구조물의경험값,기초설계및시공성과 ,현장시험자료를통하여지지력을추정할수있다 .',
]
query_embeddings = model.encode_query(queries)
document_embeddings = model.encode_document(documents)
print(query_embeddings.shape, document_embeddings.shape)
# [1, 768] [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(query_embeddings, document_embeddings)
print(similarities)
# tensor([[ 0.6528, -0.1108, -0.0670]])

Evaluation

Metrics

Information Retrieval

• Evaluated with InformationRetrievalEvaluator

Training Details

Training Dataset

Unnamed Dataset

• Size: 41,881 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: 10 tokensmean: 26.2 tokensmax: 57 tokens	min: 14 tokensmean: 259.67 tokensmax: 579 tokens

• Samples:

  sentence_0	sentence_1
  KDS 10 00 00 설계기준은 어느 나라의 표준인가요?	KDS 10 00 00설계기준 Korean Design StandardKDS 10 00 00 : 2021공통설계기준.2021년5월12일개정http://www.kcsc.re.kr
  KDS 10 00 00 설계기준은 최근에 언제 개정되었나요?	KDS 10 00 00설계기준 Korean Design StandardKDS 10 00 00 : 2021공통설계기준.2021년5월12일개정http://www.kcsc.re.kr
  KDS 10 10 00 설계총칙 문서는 어떤 분야의 설계 기준을 다루고 있나요?	공통설계기준체계KDS 10 10 00 설계총칙 `21.05

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• num_train_epochs: 13
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 8
• per_device_eval_batch_size: 8
• 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
• num_train_epochs: 13
• 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: None
• hub_always_push: False
• hub_revision: None
• 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
• 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
• liger_kernel_config: None
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs