CeLLaTe-ner-2class-pubmedbert-tapt-combData-tokenizer-adapted-spanmask
This model is a fine-tuned version of Mardiyyah/CeLLaTe-tapt-pubmedbert-tokenizer-adapted-spanmask-combinedData on the OTAR3088/CeLLaTe-ner-2class-iob_final dataset. It achieves the following results on the evaluation set:
- Loss: 0.0588
- Precision: 0.7672
- Recall: 0.7114
- Micro F1: 0.7382
- Weighted F1: 0.7384
- Macro F1: 0.7514
- Accuracy: 0.9828
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: 16
- eval_batch_size: 16
- seed: 3407
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- 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
- lr_scheduler_warmup_ratio: 0.01
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | Micro F1 | Weighted F1 | Macro F1 | Accuracy |
|---|---|---|---|---|---|---|---|---|---|
| 0.2486 | 1.0 | 263 | 0.0624 | 0.5508 | 0.6548 | 0.5984 | 0.5956 | 0.5780 | 0.9786 |
| 0.041 | 2.0 | 526 | 0.0544 | 0.7598 | 0.7132 | 0.7357 | 0.7361 | 0.7458 | 0.9837 |
| 0.026 | 3.0 | 789 | 0.0583 | 0.7672 | 0.7114 | 0.7382 | 0.7384 | 0.7514 | 0.9828 |
| 0.0191 | 4.0 | 1052 | 0.0628 | 0.7425 | 0.7336 | 0.7381 | 0.7381 | 0.7393 | 0.9834 |
| 0.0144 | 5.0 | 1315 | 0.0666 | 0.7521 | 0.7078 | 0.7292 | 0.7292 | 0.7289 | 0.9831 |
| 0.011 | 6.0 | 1578 | 0.0676 | 0.7408 | 0.7132 | 0.7267 | 0.7263 | 0.7158 | 0.9836 |
Framework versions
- Transformers 4.48.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.21.0