CeLLaTe-ner-2class-pubmedbert-tapt-tokenizer-adapted-spanmask-lr_2.16
This model is a fine-tuned version of Mardiyyah/CeLLaTe-tapt-pubmedbert-tokenizer-adapted-spanmask on the OTAR3088/CeLLaTe-ner-2class-iob_final dataset. It achieves the following results on the evaluation set:
- Loss: 0.0715
- Precision: 0.8024
- Recall: 0.7348
- Micro F1: 0.7671
- Weighted F1: 0.7668
- Macro F1: 0.7754
- Accuracy: 0.9841
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: 2.1601722686e-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: 0.02
- 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 |
|---|---|---|---|---|---|---|---|---|---|
| 0.2892 | 1.0 | 263 | 0.0692 | 0.5629 | 0.6542 | 0.6051 | 0.6075 | 0.6199 | 0.9759 |
| 0.0413 | 2.0 | 526 | 0.0550 | 0.7616 | 0.7126 | 0.7363 | 0.7363 | 0.7451 | 0.9832 |
| 0.026 | 3.0 | 789 | 0.0601 | 0.7599 | 0.7156 | 0.7371 | 0.7373 | 0.7447 | 0.9830 |
| 0.018 | 4.0 | 1052 | 0.0711 | 0.7861 | 0.7029 | 0.7422 | 0.7422 | 0.7481 | 0.9829 |
| 0.013 | 5.0 | 1315 | 0.0725 | 0.8024 | 0.7348 | 0.7671 | 0.7668 | 0.7754 | 0.9841 |
| 0.0105 | 6.0 | 1578 | 0.0767 | 0.7937 | 0.7288 | 0.7599 | 0.7597 | 0.7663 | 0.9839 |
| 0.0075 | 7.0 | 1841 | 0.0791 | 0.7409 | 0.7565 | 0.7486 | 0.7486 | 0.7490 | 0.9831 |
| 0.0065 | 8.0 | 2104 | 0.0835 | 0.7658 | 0.7511 | 0.7583 | 0.7584 | 0.7656 | 0.9836 |
| 0.0054 | 9.0 | 2367 | 0.0838 | 0.7370 | 0.7649 | 0.7507 | 0.7512 | 0.7607 | 0.9832 |
| 0.0043 | 10.0 | 2630 | 0.0954 | 0.7796 | 0.7318 | 0.7550 | 0.7550 | 0.7616 | 0.9838 |
Framework versions
- Transformers 4.48.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.21.0