bert-base-cased
This model is a fine-tuned version of google-bert/bert-base-cased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.6750
- F1 Macro: 0.9031
- F1: 0.9370
- F1 Neg: 0.8692
- Acc: 0.915
- Prec: 0.9336
- Recall: 0.9405
- Mcc: 0.8063
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 Macro | F1 | F1 Neg | Acc | Prec | Recall | Mcc |
|---|---|---|---|---|---|---|---|---|---|---|
| 0.1886 | 1.0 | 2125 | 0.3952 | 0.8938 | 0.9283 | 0.8593 | 0.905 | 0.9425 | 0.9145 | 0.7884 |
| 0.0578 | 2.0 | 4250 | 0.6750 | 0.9031 | 0.9370 | 0.8692 | 0.915 | 0.9336 | 0.9405 | 0.8063 |
| 0.0243 | 3.0 | 6375 | 0.7559 | 0.8922 | 0.9294 | 0.8550 | 0.905 | 0.9294 | 0.9294 | 0.7843 |
| 0.0084 | 4.0 | 8500 | 0.8553 | 0.9001 | 0.9353 | 0.8649 | 0.9125 | 0.9301 | 0.9405 | 0.8003 |
| 0.0131 | 5.0 | 10625 | 0.8916 | 0.8974 | 0.9333 | 0.8615 | 0.91 | 0.9299 | 0.9368 | 0.7949 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2