CeLLaTe-ner-2class-reinit_llrd-pubmedbert-tapt-tokenizer-adapted-spanmask-gazetters-lr_3.375
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: 0.1166
- Precision: 0.7920
- Recall: 0.7715
- Micro F1: 0.7816
- Weighted F1: 0.7818
- Macro F1: 0.7901
- Accuracy: 0.9844
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: 3.375114188e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 10472999
- 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: 0.05
- 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 |
|---|---|---|---|---|---|---|---|---|---|
| 0.209 | 1.0 | 1052 | 0.0581 | 0.6379 | 0.7498 | 0.6893 | 0.6901 | 0.7029 | 0.9807 |
| 0.0181 | 2.0 | 2104 | 0.0807 | 0.7924 | 0.7414 | 0.7661 | 0.7659 | 0.7770 | 0.9833 |
| 0.0066 | 3.0 | 3156 | 0.0875 | 0.7264 | 0.7823 | 0.7533 | 0.7549 | 0.7692 | 0.9825 |
| 0.0038 | 4.0 | 4208 | 0.0918 | 0.7941 | 0.7444 | 0.7685 | 0.7688 | 0.7774 | 0.9834 |
| 0.0027 | 5.0 | 5260 | 0.0979 | 0.7869 | 0.7414 | 0.7635 | 0.7635 | 0.7731 | 0.9838 |
| 0.0021 | 6.0 | 6312 | 0.1124 | 0.7767 | 0.7697 | 0.7732 | 0.7734 | 0.7797 | 0.9836 |
| 0.0014 | 7.0 | 7364 | 0.1123 | 0.7435 | 0.7529 | 0.7481 | 0.7484 | 0.7570 | 0.9829 |
| 0.0014 | 8.0 | 8416 | 0.1223 | 0.8030 | 0.7354 | 0.7677 | 0.7678 | 0.7761 | 0.9835 |
| 0.0013 | 9.0 | 9468 | 0.1165 | 0.7920 | 0.7715 | 0.7816 | 0.7818 | 0.7901 | 0.9844 |
| 0.0008 | 10.0 | 10520 | 0.1286 | 0.7992 | 0.7420 | 0.7696 | 0.7697 | 0.7782 | 0.9840 |
| 0.0006 | 11.0 | 11572 | 0.1334 | 0.7507 | 0.7841 | 0.7671 | 0.7672 | 0.7730 | 0.9832 |
| 0.0008 | 12.0 | 12624 | 0.1253 | 0.7833 | 0.7673 | 0.7752 | 0.7752 | 0.7831 | 0.9839 |
| 0.0007 | 13.0 | 13676 | 0.1335 | 0.7885 | 0.7643 | 0.7762 | 0.7763 | 0.7837 | 0.9844 |
| 0.0006 | 14.0 | 14728 | 0.1243 | 0.7595 | 0.7709 | 0.7651 | 0.7654 | 0.7756 | 0.9834 |
Framework versions
- Transformers 4.48.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.21.0