Intellexus/Bi-Tib-mbert-v4

SentenceTransformer based on Shailu1492/tibetan-mbert-v1-consecutive-segments

This is a sentence-transformers model finetuned from Shailu1492/tibetan-mbert-v1-consecutive-segments. 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: Shailu1492/tibetan-mbert-v1-consecutive-segments
• Maximum Sequence Length: 512 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': 512, 'do_lower_case': False, 'architecture': 'BertModel'})
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': True, '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("sentence_transformers_model_id")
# Run inference
sentences = [
    "bden pa'i gsung gis rab tu bka' stsal na",
    "bden pa gsung gis rab tu bka' stsal na",
    "byang chub sems kyi bsod nams gang // gal te de la gzugs mchis na// nam mkha'i khams 'di kun gang ste// de ni de bas lhag par 'gyur//",
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.9909, 0.7003],
#         [0.9909, 1.0000, 0.7065],
#         [0.7003, 0.7065, 1.0000]])

Evaluation

Metrics

Semantic Similarity

• Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.9507
spearman_cosine	0.9535

Training Details

Training Dataset

Unnamed Dataset

• Size: 1,000 training samples
• Columns: text1, text2, and label
• Approximate statistics based on the first 1000 samples:

  	text1	text2	label
  type	string	string	float
  details	min: 8 tokens mean: 30.7 tokens max: 190 tokens	min: 6 tokens mean: 30.97 tokens max: 226 tokens	min: 0.0 mean: 0.5 max: 1.0

• Samples:

  text1	text2	label
  brdar brtags chos nyid thob pa ste// sdom la gnas pa rnams la yod//	sgra rnams kyis ni brdar btags ston// de ni tha snyad ched du byas//	0.3125
  zhe sdang gi rang bzhin gsal ba ste/	gti mug gi rang bzhin mi rtog par shes/	0.125
  gsang ba'i dbang dang shes rab ye shes kyi dbang dang bzhi pa'i dbang zhu bar bya'o//	gsang ba'i dbang ni ngag dag par byed pa'o// shes rab ye shes kyi dbang ni yid dag par byed pa'o//	0.5

• Loss: CoSENTLoss with these parameters:

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

Evaluation Dataset

Unnamed Dataset

• Size: 1,000 evaluation samples
• Columns: text1, text2, and label
• Approximate statistics based on the first 1000 samples:

  	text1	text2	label
  type	string	string	float
  details	min: 8 tokens mean: 30.7 tokens max: 190 tokens	min: 6 tokens mean: 30.97 tokens max: 226 tokens	min: 0.0 mean: 0.5 max: 1.0

• Samples:

  text1	text2	label
  brdar brtags chos nyid thob pa ste// sdom la gnas pa rnams la yod//	sgra rnams kyis ni brdar btags ston// de ni tha snyad ched du byas//	0.3125
  zhe sdang gi rang bzhin gsal ba ste/	gti mug gi rang bzhin mi rtog par shes/	0.125
  gsang ba'i dbang dang shes rab ye shes kyi dbang dang bzhi pa'i dbang zhu bar bya'o//	gsang ba'i dbang ni ngag dag par byed pa'o// shes rab ye shes kyi dbang ni yid dag par byed pa'o//	0.5

• Loss: CoSENTLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• per_device_train_batch_size: 32
• gradient_accumulation_steps: 16
• learning_rate: 2e-05
• weight_decay: 0.1
• num_train_epochs: 7
• lr_scheduler_type: reduce_lr_on_plateau
• warmup_ratio: 0.1
• warmup_steps: 0.1
• bf16: True
• dataloader_drop_last: True
• load_best_model_at_end: True

All Hyperparameters

Click to expand

• do_predict: False
• eval_strategy: epoch
• prediction_loss_only: True
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 8
• gradient_accumulation_steps: 16
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 2e-05
• weight_decay: 0.1
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 7
• max_steps: -1
• lr_scheduler_type: reduce_lr_on_plateau
• lr_scheduler_kwargs: None
• warmup_ratio: 0.1
• warmup_steps: 0.1
• log_level: passive
• log_level_replica: warning
• log_on_each_node: True
• logging_nan_inf_filter: True
• enable_jit_checkpoint: False
• save_on_each_node: False
• save_only_model: False
• restore_callback_states_from_checkpoint: False
• use_cpu: False
• seed: 42
• data_seed: None
• bf16: True
• fp16: False
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: None
• local_rank: -1
• ddp_backend: None
• debug: []
• dataloader_drop_last: True
• dataloader_num_workers: 0
• dataloader_prefetch_factor: None
• disable_tqdm: False
• remove_unused_columns: True
• label_names: None
• load_best_model_at_end: True
• ignore_data_skip: False
• fsdp: []
• fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
• accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
• parallelism_config: None
• deepspeed: None
• label_smoothing_factor: 0.0
• optim: adamw_torch_fused
• optim_args: None
• group_by_length: False
• length_column_name: length
• project: huggingface
• trackio_space_id: trackio
• 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
• 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_for_metrics: []
• eval_do_concat_batches: True
• auto_find_batch_size: False
• full_determinism: False
• ddp_timeout: 1800
• torch_compile: False
• torch_compile_backend: None
• torch_compile_mode: None
• include_num_input_tokens_seen: no
• 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: True
• use_cache: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	Validation Loss	spearman_cosine
1.0	2	6.0687	2.5858	0.8629
2.0	4	5.5189	2.5479	0.8886
3.0	6	5.4863	2.4755	0.9129
4.0	8	5.3785	2.3440	0.9284
5.0	10	5.2816	2.2080	0.9355
6.0	12	5.2027	2.0943	0.9432
7.0	14	5.0971	1.9869	0.9535

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 5.2.2
• Transformers: 5.1.0
• PyTorch: 2.10.0+cu130
• Accelerate: 1.12.0
• Datasets: 4.4.1
• Tokenizers: 0.22.2

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",
}

CoSENTLoss

@article{10531646,
    author={Huang, Xiang and Peng, Hao and Zou, Dongcheng and Liu, Zhiwei and Li, Jianxin and Liu, Kay and Wu, Jia and Su, Jianlin and Yu, Philip S.},
    journal={IEEE/ACM Transactions on Audio, Speech, and Language Processing},
    title={CoSENT: Consistent Sentence Embedding via Similarity Ranking},
    year={2024},
    doi={10.1109/TASLP.2024.3402087}
}