curiousabhinav/finetuned-bge-base-en

SentenceTransformer based on BAAI/bge-base-en

This is a sentence-transformers model finetuned from BAAI/bge-base-en. 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: BAAI/bge-base-en
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 tokens
• 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': True}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, '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): 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("curiousabhinav/finetuned-bge-base-en")
# Run inference
sentences = [
    '\nName : BlueHorizon Retreats\nCategory: Event Planning Services, Executive Travel Arrangements\nDepartment: Executive\nLocation: Maui, Hawaii\nAmount: 2356.5\nCard: Annual Leadership Summit\nTrip Name: Executive Team Building Retreat\n',
    '\nName : InteractivoPlus\nCategory: Communication Platforms, Subscription Services\nDepartment: Customer Success\nLocation: Barcelona, Spain\nAmount: 1157.49\nCard: Enhanced Communication Channel\nTrip Name: unknown\n',
    '\nName : TransDia Solutions\nCategory: Automotive Services, Logistics Consulting\nDepartment: Sales\nLocation: Bordeaux, France\nAmount: 1326.47\nCard: Client Distribution Coordination\nTrip Name: Quarterly Regional Visits\n',
]
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

Triplet

• Dataset: bge-base-en-train
• Evaluated with TripletEvaluator

Metric	Value
cosine_accuracy	0.8606
dot_accuracy	0.1394
manhattan_accuracy	0.8654
euclidean_accuracy	0.8606
max_accuracy	0.8654

Triplet

• Dataset: bge-base-en-eval
• Evaluated with TripletEvaluator

Metric	Value
cosine_accuracy	0.9545
dot_accuracy	0.0455
manhattan_accuracy	0.9545
euclidean_accuracy	0.9545
max_accuracy	0.9545

Training Details

Training Dataset

Unnamed Dataset

• Size: 208 training samples
• Columns: sentence and label
• Approximate statistics based on the first 208 samples:

  	sentence	label
  type	string	int
  details	min: 33 tokens mean: 39.81 tokens max: 49 tokens	0: ~3.85% 1: ~3.37% 2: ~3.85% 3: ~2.40% 4: ~5.29% 5: ~4.33% 6: ~4.33% 7: ~3.37% 8: ~3.85% 9: ~4.33% 10: ~3.37% 11: ~3.85% 12: ~2.40% 13: ~5.29% 14: ~3.37% 15: ~5.77% 16: ~4.33% 17: ~2.40% 18: ~2.88% 19: ~3.37% 20: ~3.85% 21: ~4.33% 22: ~2.88% 23: ~4.33% 24: ~4.81% 25: ~1.92% 26: ~1.92%

• Samples:

  sentence	label
  Name : Transcend Category: Upskilling Department: Human Resource Location: London, UK Amount: 859.47 Card: Technology Skills Enhancement Trip Name: unknown	0
  Name : Ayden Category: Financial Software Department: Finance Location: Berlin, DE Amount: 1273.45 Card: Enterprise Technology Services Trip Name: unknown	1
  Name : Urban Sphere Category: Utilities Management, Facility Services Department: Office Administration Location: New York, NY Amount: 937.32 Card: Monthly Operations Budget Trip Name: unknown	2

• Loss: BatchSemiHardTripletLoss

Evaluation Dataset

Unnamed Dataset

• Size: 66 evaluation samples
• Columns: sentence and label
• Approximate statistics based on the first 66 samples:

  	sentence	label
  type	string	int
  details	min: 35 tokens mean: 40.83 tokens max: 46 tokens	0: ~3.03% 1: ~1.52% 2: ~1.52% 4: ~3.03% 5: ~4.55% 6: ~1.52% 7: ~9.09% 8: ~6.06% 9: ~3.03% 10: ~4.55% 11: ~3.03% 12: ~3.03% 13: ~3.03% 14: ~3.03% 15: ~4.55% 16: ~4.55% 18: ~3.03% 19: ~3.03% 20: ~7.58% 21: ~1.52% 22: ~6.06% 23: ~4.55% 24: ~4.55% 25: ~3.03% 26: ~7.58%

• Samples:

  sentence	label
  Name : Aperio Global Insights Category: Strategic Business Consulting, Data Analytics Services Department: Finance Location: Chicago, IL Amount: 3456.78 Card: Global Market Expansion Evaluation Trip Name: unknown	5
  Name : InfiniTech Systems Category: Technical Hardware Management, Software Solutions Department: IT Operations Location: New York, USA Amount: 1099.47 Card: Integrated Support Package Trip Name: unknown	12
  Name : Mosaic Labour Mitachi Category: Labour Consultancy, Professional Workforce Assessment Department: Human Resource Location: Tokyo, Japan Amount: 2985.65 Card: Global Workforce Optimization Trip Name: unknown	5

• Loss: BatchSemiHardTripletLoss

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• learning_rate: 2e-05
• num_train_epochs: 5
• warmup_ratio: 0.1
• dataloader_pin_memory: False
• 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: 16
• per_device_eval_batch_size: 16
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• learning_rate: 2e-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: 5
• 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: 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: False
• 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: False
• hub_always_push: False
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• 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
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	bge-base-en-eval_max_accuracy	bge-base-en-train_max_accuracy
0	0	-	0.8654
5.0	65	0.9545	-

Framework Versions

• Python: 3.12.7
• Sentence Transformers: 3.1.1
• Transformers: 4.42.4
• PyTorch: 2.7.0
• Accelerate: 1.3.0
• Datasets: 3.6.0
• Tokenizers: 0.19.1

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",
}

BatchSemiHardTripletLoss

@misc{hermans2017defense,
    title={In Defense of the Triplet Loss for Person Re-Identification},
    author={Alexander Hermans and Lucas Beyer and Bastian Leibe},
    year={2017},
    eprint={1703.07737},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}