nicolauduran45 / e5-finetuned-entity-linking

SentenceTransformer based on intfloat/multilingual-e5-base

This is a sentence-transformers model finetuned from intfloat/multilingual-e5-base . 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: intfloat/multilingual-e5-base
• 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': 'XLMRobertaModel'})
  (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): 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("nicolauduran45/e5-finetuned-entity-linking")
# Run inference
sentences = [
    'query: liquefied petroleum gas carriers. Context: Using fuel consumption as a main input, Psaraftis and Kontovas [25] presented an analysis on emissions of the world fleet for major ship types, i.e., bulk carriers, crude oil tankers, container vessels, product or chemical carriers, liquefied natural gas (liquefied natural gas) carriers, liquefied petroleum gas carriers, reefer vessels, Ro-Ro vessels, and general cargo ships.',
    'passage: Gas Carrier [SEP] A tanker equipped with specialized pressurized or refrigerated tanks for transporting liquefied natural gas (LNG), liquefied petroleum gas (LPG), or other gaseous products. [SEP] gas tanker',
    'passage: Offshore Supply Vessel [SEP] A ship designed to transport supplies, equipment, and personnel to offshore oil and gas platforms, wind farms, and other marine installations. [SEP] PSV ,  platform supply vessel ,  offshore supply ship ,  supply ship ,  offshore supply vessel ,  platform supply ship',
]
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, 0.8028, 0.2250],
#         [0.8028, 1.0000, 0.2569],
#         [0.2250, 0.2569, 1.0000]])

Training Details

Training Dataset

Unnamed Dataset

• Size: 1,260 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: 16 tokens mean: 63.57 tokens max: 338 tokens	min: 15 tokens mean: 77.89 tokens max: 368 tokens

• Samples:

  sentence_0	sentence_1
  query: tugboat. Context: Sharing deployment information: Information sharing between the pilot, tugboat captain and boatmen crew with their planning departments regarding the deployment information such as meeting point with the vessel, or estimated start and completion time of services.	passage: Tugboat [SEP] A powerful, maneuverable vessel used to push or tow other ships in harbors, through canals, or at sea, assisting with docking and navigation in confined waters. [SEP] tugboat , tug boat , tug , water tractor
  query: fluorescence microscope. Context: staining of the striatum and the mean fluorescence density was performed through a fluorescence microscope, green fluorescence intensity represents the expression level of tyrosine hydroxylase, the results showed that the green fluorescence in the experimental group was slightly lower than that in the Sham group (Fig.	passage: fluorescence microscopy [SEP] Fluorescence microscopy comprises any type of microscopy where the specimen can be made to fluoresce (emit energy as visible light), typically by illuminating it with light of specific wavelengths. [SEP] Microscopy, Fluorescence
  query: of liquefied natural gas carriers. Context: Transverse acceleration: Longitudinal acceleration: where: (2) Forced displacement of hull deformation The topic of this study is the piping system installed on the deck of liquefied natural gas carriers.	passage: Gas Carrier [SEP] A tanker equipped with specialized pressurized or refrigerated tanks for transporting liquefied natural gas (LNG), liquefied petroleum gas (LPG), or other gaseous products. [SEP] gas tanker

• Loss: MultipleNegativesRankingLoss with these parameters:

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

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size : 32
• per_device_eval_batch_size : 32
• fp16 : True
• multi_dataset_batch_sampler : round_robin

All Hyperparameters

Framework Versions

• Python: 3.12.12
• Sentence Transformers: 5.2.2
• Transformers: 5.0.0
• PyTorch: 2.9.0+cu128
• Accelerate: 1.12.0
• Datasets: 4.0.0
• 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",
}

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}
}