README.md

---
base_model: sentence-transformers/paraphrase-multilingual-mpnet-base-v2
language:
- en
- es
- de
- zh
- mul
- multilingual
library_name: sentence-transformers
license: mit
pipeline_tag: feature-extraction
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:21123868
- loss:CachedMultipleNegativesRankingLoss
widget:
- source_sentence: 系统管理员技术员——TS/SCI级别并拥有多项式验证
  sentences:
  - support development of annual budget, create a financial report, report analysis
    results, Microsoft Access, accounting, use presentation software, interpret financial
    statements, synthesise financial information, develop vaccines, handle financial
    overviews of the store, produce statistical financial records, develop financial
    statistics reports, explain accounting records, financial analysis, SAP R3, represent
    the company, examine budgets, prepare presentation material, use spreadsheets
    software, forecast account metrics, meet deadlines, prepare financial projections,
    manage budgets, exercise self-control, financial statements
  - ensure cross-department cooperation, establish customer rapport, improve business
    processes, manage technical security systems, handle incidents, maintain ICT system,
    explain characteristics of computer peripheral equipment, gather technical information,
    collaborate in company's daily operations , apply change management, maintain
    technical equipment, communicate with customers, solve technical problems, perform
    ICT troubleshooting, use ICT equipment in maintenance activities, manage major
    incidents, build business relationships, computer engineering, perform software
    recovery testing, identify process improvements, maintain relationship with customers,
    carry out project activities, collaborate in the development of marketing strategies,
    computer technology, technical terminology
  - utilise machine learning, cloud technologies, develop predictive models, assess
    sportive performance, formulate findings , principles of artificial intelligence,
    perform business research, communicate with stakeholders, computer engineering,
    build predictive models, computer science, develop automated software tests, analyse
    business objectives, Agile development, cloud monitoring and reporting, provide
    written content, obtain relevant licenses, design prototypes, machine learning,
    e-learning software infrastructure, analyse education system, disseminate results
    to the scientific community, learning technologies, ML (computer programming),
    task algorithmisation
- source_sentence: 安全运营官
  sentences:
  - deliver outstanding service, manage carriers, direct customers to merchandise,
    improve customer interaction, manage time, support managers, assist customers,
    process customer orders, manage customer service, satisfy customers, guarantee
    customer satisfaction, respond to customers' inquiries
  - manage several projects, implement operational business plans, identify improvement
    actions, develop strategy to solve problems, manage website, carry out project
    activities, follow reporting procedures, supervise site maintenance, adjust priorities,
    schedule shifts, conduct public presentations, motivate others, manage operational
    budgets, report to the team leader, encourage teams for continuous improvement,
    lead the sustainability reporting process, implement sustainable procurement,
    show an exemplary leading role in an organisation, manage manufacturing facilities,
    develop training programmes, develop production line, supply chain management,
    leadership principles, lead a team, coaching techniques
  - provide emergency supplies, provide first aid, liaise with security authorities,
    apply medical first aid in case of emergency, regulate traffic, train security
    officers, maintain physical fitness, provide protective escort, ensure public
    safety and security, ensure inspections of facilities, work in inclement conditions,
    follow procedures in the event of an alarm, set safety and security standards,
    comply with the principles of self-defence, present reports, maintain facility
    security systems, conduct security screenings, types of evaluation , monitor security
    measures, office equipment, escort pedestrians across streets, advise on security
    staff selection, wear appropriate protective gear, work in outdoor conditions,
    assist emergency services
- source_sentence: Empleado de control de COVID
  sentences:
  - maintain records of clients' prescriptions, assist people in contaminated areas,
    label samples, maintain museum records, apply social distancing protocols, collect
    biological samples from patients, infection control, label medical laboratory
    samples, disinfect surfaces, maintain customer records, ensure health and safety
    of staff, personal protective equipment, remove contaminated materials, store
    contaminated materials, prepare prescription labels, use personal protection equipment
  - promote organisational communication, provide legal advice, human resource management,
    company policies, perform customer management, business processes, ensure compliance
    with legal requirements, develop communications strategies, enforce company values,
    develop outreach training plans, use consulting techniques, develop employment
    policies, human resources department processes, personnel management, identify
    training needs, participate in health personnel training, health and safety in
    the workplace, lead police investigations, ensure compliance with policies, prepare
    compliance documents, perform internal investigations, develop employee retention
    programs, develop corporate training programmes, customer relationship management,
    manage localisation
  - perform escalation procedure, imprint visionary aspirations into the business
    management, observe confidentiality, impart business plans to collaborators, lead
    a team, human resources department processes, respect confidentiality obligations,
    hire human resources, manage commercial risks, develop business plans, communicate
    with stakeholders, maintain relationship with customers, manage several projects,
    provide improvement strategies, manage technical security systems, knowledge management,
    risk management, develop program ideas, perform project management, project management,
    cope with uncertainty, address identified risks, provide performance feedback,\
    information confidentiality, track key performance indicators
- source_sentence: Aerie - Brand Ambassador (Sales Associate) - US
  sentences:
  - lay bricks, provide first aid, enforce park rules, conflict management, give swimming
    lessons, assist in performing physical exercises, perform park safety inspections,
    assist in the movement of heavy loads, lead a team, first aid, supervise pool
    activities, swim, coach staff for running the performance, show an exemplary leading
    role in an organisation, teach public speaking principles, collaborate with coaching
    team, supervise work, calculate stairs rise and run, calculate compensation payments,
    manage a team, information confidentiality
  - react to events in time-critical environments, operate in a specific field of
    nursing care, clinical science, promote healthy fitness environment, lead others,
    comply with legislation related to health care, maintain a safe, hygienic and
    secure working environment, provide healthcare services to patients in specialised
    medicine, write English, conduct physical examinations, leadership principles,
    use clinical assessment techniques, apply context specific clinical competences,
    conduct health related research, conceptualise healthcare user’s needs, assessment
    processes, communicate in healthcare, provide professional care in nursing, nursing
    science, promote health and safety, implement policy in healthcare practices,
    engage with stakeholders, identify problems, respond to changing situations in
    health care, perform resource planning
  - ensure the privacy of guests, provide customised products, company policies, exude
    enthusiasm during the action sessions, provide customer guidance on product selection,
    collect briefing regarding products, perform multiple tasks at the same time,\
    create solutions to problems, respond to visitor complaints
- source_sentence: 医师——危重症护理——重症监护专家——项目医务总监
  sentences:
  - handle incidents, provide technical documentation, coordinate operational activities,
    ensure information security, work in teams, manage manufacturing documentation,
    project configuration management, operate call distribution system, maintain computer
    hardware, apply change management, manage aircraft support systems, perform escalation
    procedure, manage production changeovers, maintenance operations, call-centre
    technologies, manage service contracts in the drilling industry, encourage teambuilding,
    manage major incidents, resolve equipment malfunctions, work independently, think
    analytically, manage maintenance operations, maintain plan for continuity of operations
  - develop recycling programs, receive actors' resumes, work in cold environments,
    perform cleaning duties, operate floor cleaning equipment, operate forklift
  - perform technical tasks with great care, supervise medical residents, manage a
    multidisciplinary team involved in patient care, administrative tasks in a medical
    environment, demonstrate technical skills during neurological surgery, apply problem
    solving in social service, intensive care medicine, provide comprehensive care
    for patients with surgical conditions, work in teams, solve problems
co2_eq_emissions:
  emissions: 717.3535184611766
  energy_consumed: 1.9440474755045436
  source: codecarbon
  training_type: fine-tuning
  on_cloud: true
  cpu_model: Intel(R) Xeon(R) CPU @ 2.20GHz
  ram_total_size: 83.47684860229492
  hours_used: 5.34
  hardware_used: 1 x NVIDIA A100-SXM4-40GB
---

# SentenceTransformer based on sentence-transformers/paraphrase-multilingual-mpnet-base-v2

This is a [sentence-transformers](https://www.SBERT.net) model specifically trained for job title matching and similarity. It's finetuned from [sentence-transformers/paraphrase-multilingual-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-multilingual-mpnet-base-v2) on a large dataset of job titles and their associated skills/requirements across multiple languages. The model maps English, Spanish, German and Chinese job titles and descriptions to a 1024-dimensional dense vector space and can be used for semantic job title matching, job similarity search, and related HR/recruitment tasks.

The model was presented in the paper [Multilingual JobBERT for Cross-Lingual Job Title Matching](https://huggingface.co/papers/2507.21609).

## Model Details

### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [sentence-transformers/paraphrase-multilingual-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-multilingual-mpnet-base-v2) <!-- at revision 84fccfe766bcfd679e39efefe4ebf45af190ad2d -->
- **Maximum Sequence Length:** 64 tokens
- **Output Dimensionality:** 1024 dimensions
- **Similarity Function:** Cosine Similarity
- **Training Dataset:** 4 x 5.2M high-quality job title - skills pairs in English, Spanish, German and Chinese

### Model Sources

- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)

### Full Model Architecture

```
SentenceTransformer(
  (0): Transformer({'max_seq_length': 64, 'do_lower_case': False}) with Transformer model: 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): Asym(
    (anchor-0): Dense({'in_features': 768, 'out_features': 1024, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
    (positive-0): Dense({'in_features': 768, 'out_features': 1024, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
  )
)
```

## Usage

### Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

```bash
pip install -U sentence-transformers
```

Then you can load and use the model with the following code:
```python
import torch
import numpy as np
from tqdm.auto import tqdm
from sentence_transformers import SentenceTransformer
from sentence_transformers.util import batch_to_device, cos_sim

# Load the model
model = SentenceTransformer("TechWolf/JobBERT-v3")

def encode_batch(jobbert_model, texts):
    features = jobbert_model.tokenize(texts)
    features = batch_to_device(features, jobbert_model.device)
    features["text_keys"] = ["anchor"]
    with torch.no_grad():
        out_features = jobbert_model.forward(features)
    return out_features["sentence_embedding"].cpu().numpy()

def encode(jobbert_model, texts, batch_size: int = 8):
    # Sort texts by length and keep track of original indices
    sorted_indices = np.argsort([len(text) for text in texts])
    sorted_texts = [texts[i] for i in sorted_indices]
    
    embeddings = []
    
    # Encode in batches
    for i in tqdm(range(0, len(sorted_texts), batch_size)):
        batch = sorted_texts[i:i+batch_size]
        embeddings.append(encode_batch(jobbert_model, batch))
    
    # Concatenate embeddings and reorder to original indices
    sorted_embeddings = np.concatenate(embeddings)
    original_order = np.argsort(sorted_indices)
    return sorted_embeddings[original_order]

# Example usage
job_titles = [
    'Software Engineer',
    '高级软件开发人员',  # senior software developer
    'Produktmanager',  # product manager
    'Científica de datos'  # data scientist
]

# Get embeddings
embeddings = encode(model, job_titles)

# Calculate cosine similarity matrix
similarities = cos_sim(embeddings, embeddings)
print(similarities)
```

The output will be a similarity matrix where each value represents the cosine similarity between two job titles:

```
tensor([[1.0000, 0.8087, 0.4673, 0.5669],
        [0.8087, 1.0000, 0.4428, 0.4968],
        [0.4673, 0.4428, 1.0000, 0.4292],
        [0.5669, 0.4968, 0.4292, 1.0000]])
```


<!--
### Direct Usage (Transformers)

<details><summary>Click to see the direct usage in Transformers</summary>

</details>
-->

<!--
### Downstream Usage (Sentence Transformers)

You can finetune this model on your own dataset.

<details><summary>Click to expand</summary>

</details>
-->

<!--
### Out-of-Scope Use

*List how the model may foreseeably be misused and address what users ought not to do with the model.*
-->

<!--
## Bias, Risks and Limitations

*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
-->

<!--
### Recommendations

*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
-->

## Training Details

### Training Dataset

* Size: 21,123,868 training samples
* Columns: <code>anchor</code> and <code>positive</code>
* Approximate statistics based on the first 1000 samples:
  |         | anchor                                                                            | positive                                                                           |
  |:--------|:----------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|
  | type    | string                                                                            | string                                                                             |
  | details | <ul><li>min: 4 tokens</li><li>mean: 10.56 tokens</li><li>max: 38 tokens</li></ul> | <ul><li>min: 19 tokens</li><li>mean: 61.08 tokens</li><li>max: 64 tokens</li></ul> |
* Samples:
  | anchor                                         | positive                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    |
  |:-----------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
  | <code>通信与培训专员</code>                           | <code>deliver online training, liaise with educational support staff, interact with an audience, construct individual learning plans, lead a team, develop corporate training programmes, learning technologies, communication, identify with the company's goals, address an audience, learning management systems, use presentation software, motivate others, provide learning support, engage with stakeholders, identify skills gaps, meet expectations of target audience, develop training programmes</code>                                                                                                                                         |
  | <code>Associate Infrastructure Engineer</code> | <code>create solutions to problems, design user interface, cloud technologies, use databases, automate cloud tasks, keep up-to-date to computer trends, work in teams, use object-oriented programming, keep updated on innovations in various business fields, design principles, Angular, adapt to changing situations, JavaScript, Agile development, manage stable, Swift (computer programming), keep up-to-date to design industry trends, monitor technology trends, web programming, provide mentorship, advise on efficiency improvements, adapt to change, JavaScript Framework, database management systems, stimulate creative processes</code> |
  | <code>客户顾问/出纳</code>                           | <code>customer service, handle financial transactions, adapt to changing situations, have computer literacy, manage cash desk, attend to detail, provide customer guidance on product selection, perform multiple tasks at the same time, carry out financial transactions, provide membership service, manage accounts, adapt to change, identify customer's needs, solve problems</code>                                                                                                                                                                                                                                                                  |
* Loss: [<code>CachedMultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#cachedmultiplenegativesrankingloss) with these parameters:
  ```json
  {
      "scale": 20.0,
      "similarity_fct": "cos_sim",
      "mini_batch_size": 512
  }
  ```

### Training Hyperparameters
#### Non-Default Hyperparameters

- `overwrite_output_dir`: True
- `per_device_train_batch_size`: 2048
- `per_device_eval_batch_size`: 2048
- `num_train_epochs`: 1
- `fp16`: True

#### All Hyperparameters
<details><summary>Click to expand</summary>

- `overwrite_output_dir`: True
- `do_predict`: False
- `eval_strategy`: no
- `prediction_loss_only`: True
- `per_device_train_batch_size`: 2048
- `per_device_eval_batch_size`: 2048
- `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`: 1.0
- `num_train_epochs`: 1
- `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`: True
- `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
- `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`: batch_sampler
- `multi_dataset_batch_sampler`: proportional

</details>

### Training Logs
| Epoch  | Step  | Training Loss |
|:------:|:-----:|:-------------:|
| 0.0485 | 500   | 3.89          |
| 0.0969 | 1000  | 3.373         |
| 0.1454 | 1500  | 3.1715        |
| 0.1939 | 2000  | 3.0414        |
| 0.2424 | 2500  | 2.9462        |
| 0.2908 | 3000  | 2.8691        |
| 0.3393 | 3500  | 2.8048        |
| 0.3878 | 4000  | 2.7501        |
| 0.4363 | 4500  | 2.7026        |
| 0.4847 | 5000  | 2.6601        |
| 0.5332 | 5500  | 2.6247        |
| 0.5817 | 6000  | 2.5951        |
| 0.6302 | 6500  | 2.5692        |
| 0.6786 | 7000  | 2.5447        |
| 0.7271 | 7500  | 2.5221        |
| 0.7756 | 8000  | 2.5026        |
| 0.8240 | 8500  | 2.4912        |
| 0.8725 | 9000  | 2.4732        |
| 0.9210 | 9500  | 2.4608        |
| 0.9695 | 10000 | 2.4548        |


### Environmental Impact
Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon).
- **Energy Consumed**: 1.944 kWh
- **Carbon Emitted**: 0.717 kg of CO2
- **Hours Used**: 5.34 hours

### Training Hardware
- **On Cloud**: Yes
- **GPU Model**: 1 x NVIDIA A100-SXM4-40GB
- **CPU Model**: Intel(R) Xeon(R) CPU @ 2.20GHz
- **RAM Size**: 83.48 GB

### Framework Versions
- Python: 3.10.16
- Sentence Transformers: 4.1.0
- Transformers: 4.48.3
- PyTorch: 2.6.0+cu126
- Accelerate: 1.3.0
- Datasets: 3.5.1
- Tokenizers: 0.21.0

## Citation

### BibTeX

#### JobBERT-v3 Paper
```bibtex
@misc{decorte2025multilingualjobbertcrosslingualjob,
      title={Multilingual JobBERT for Cross-Lingual Job Title Matching}, 
      author={Jens-Joris Decorte and Matthias De Lange and Jeroen Van Hautte},
      year={2025},
      eprint={2507.21609},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2507.21609}, 
}
```

#### Sentence Transformers
```bibtex
@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",
}
```

#### CachedMultipleNegativesRankingLoss
```bibtex
@misc{gao2021scaling,
    title={Scaling Deep Contrastive Learning Batch Size under Memory Limited Setup},
    author={Luyu Gao and Yunyi Zhang and Jiawei Han and Jamie Callan},
    year={2021},
    eprint={2101.06983},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}
```

<!--
## Glossary

*Clearly define terms in order to be accessible across audiences.*
-->

<!--
## Model Card Authors

*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
-->

<!--
## Model Card Contact

*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
-->