resume-ner-xlm
This model is a fine-tuned version of xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1579
- Certification: {'precision': 0.937007874015748, 'recall': 0.85, 'f1': 0.8913857677902622, 'number': 140}
- Domain Skill: {'precision': 0.820036092510804, 'recall': 0.8150237179335897, 'f1': 0.8175222223537347, 'number': 42373}
- Language: {'precision': 0.9469964664310954, 'recall': 0.9816849816849816, 'f1': 0.9640287769784173, 'number': 273}
- Soft Skill: {'precision': 0.8625, 'recall': 0.9409090909090909, 'f1': 0.9000000000000001, 'number': 220}
- Tech Skill: {'precision': 0.8503740648379052, 'recall': 0.8938401048492791, 'f1': 0.8715654952076678, 'number': 763}
- Overall Precision: 0.8220
- Overall Recall: 0.8182
- Overall F1: 0.8201
- Overall Accuracy: 0.9433
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Certification | Domain Skill | Language | Soft Skill | Tech Skill | Overall Precision | Overall Recall | Overall F1 | Overall Accuracy |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.2056 | 1.0 | 566 | 0.2044 | {'precision': 0.940677966101695, 'recall': 0.7928571428571428, 'f1': 0.8604651162790697, 'number': 140} | {'precision': 0.7562669888251283, 'recall': 0.7682250489698629, 'f1': 0.7621991196028847, 'number': 42373} | {'precision': 0.9434628975265018, 'recall': 0.978021978021978, 'f1': 0.9604316546762589, 'number': 273} | {'precision': 0.8817204301075269, 'recall': 0.37272727272727274, 'f1': 0.5239616613418531, 'number': 220} | {'precision': 0.6643356643356644, 'recall': 0.747051114023591, 'f1': 0.7032695866748919, 'number': 763} | 0.7564 | 0.7673 | 0.7618 | 0.9245 |
| 0.1506 | 2.0 | 1132 | 0.1642 | {'precision': 0.937007874015748, 'recall': 0.85, 'f1': 0.8913857677902622, 'number': 140} | {'precision': 0.8205880913939038, 'recall': 0.7916361834186865, 'f1': 0.8058521819557724, 'number': 42373} | {'precision': 0.9535714285714286, 'recall': 0.978021978021978, 'f1': 0.9656419529837251, 'number': 273} | {'precision': 0.8481012658227848, 'recall': 0.9136363636363637, 'f1': 0.8796498905908097, 'number': 220} | {'precision': 0.8265179677819083, 'recall': 0.8741808650065531, 'f1': 0.8496815286624204, 'number': 763} | 0.8221 | 0.7950 | 0.8083 | 0.9404 |
| 0.1498 | 3.0 | 1698 | 0.1579 | {'precision': 0.937007874015748, 'recall': 0.85, 'f1': 0.8913857677902622, 'number': 140} | {'precision': 0.820036092510804, 'recall': 0.8150237179335897, 'f1': 0.8175222223537347, 'number': 42373} | {'precision': 0.9469964664310954, 'recall': 0.9816849816849816, 'f1': 0.9640287769784173, 'number': 273} | {'precision': 0.8625, 'recall': 0.9409090909090909, 'f1': 0.9000000000000001, 'number': 220} | {'precision': 0.8503740648379052, 'recall': 0.8938401048492791, 'f1': 0.8715654952076678, 'number': 763} | 0.8220 | 0.8182 | 0.8201 | 0.9433 |
Framework versions
- Transformers 5.12.1
- Pytorch 2.5.1+cu121
- Datasets 5.0.0
- Tokenizers 0.22.2