CeLLaTe-ner-2class-pubmedbert-tapt-combData-tokenizer-adapted-wwmask
This model is a fine-tuned version of Mardiyyah/CeLLaTe-tapt-pubmedbert-tokenizer-adapted-wwmask-combinedData on the OTAR3088/CeLLaTe-ner-2class-iob_final dataset. It achieves the following results on the evaluation set:
- Loss: 0.0655
- Precision: 0.7687
- Recall: 0.7492
- Micro F1: 0.7588
- Weighted F1: 0.7590
- Macro F1: 0.7636
- Accuracy: 0.9842
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 |
|---|---|---|---|---|---|---|---|---|---|
| 0.2553 | 1.0 | 263 | 0.0633 | 0.5902 | 0.6236 | 0.6064 | 0.6057 | 0.5988 | 0.9792 |
| 0.0427 | 2.0 | 526 | 0.0594 | 0.6989 | 0.6657 | 0.6819 | 0.6829 | 0.6965 | 0.9811 |
| 0.0265 | 3.0 | 789 | 0.0603 | 0.7281 | 0.7246 | 0.7263 | 0.7270 | 0.7438 | 0.9831 |
| 0.0197 | 4.0 | 1052 | 0.0630 | 0.7325 | 0.7342 | 0.7333 | 0.7341 | 0.7445 | 0.9836 |
| 0.0145 | 5.0 | 1315 | 0.0775 | 0.7895 | 0.6723 | 0.7262 | 0.7260 | 0.7341 | 0.9822 |
| 0.0121 | 6.0 | 1578 | 0.0651 | 0.7687 | 0.7492 | 0.7588 | 0.7590 | 0.7636 | 0.9842 |
| 0.0092 | 7.0 | 1841 | 0.0725 | 0.7509 | 0.7450 | 0.7480 | 0.7481 | 0.7521 | 0.9842 |
| 0.0072 | 8.0 | 2104 | 0.0704 | 0.7529 | 0.7402 | 0.7465 | 0.7466 | 0.7496 | 0.9842 |
| 0.0061 | 9.0 | 2367 | 0.0805 | 0.7371 | 0.7468 | 0.7419 | 0.7426 | 0.7535 | 0.9834 |
Framework versions
- Transformers 4.48.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.21.0