bert-base-cased-finetuned-ner
This model is a fine-tuned version of google-bert/bert-base-cased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0948
- Precision: 0.8623
- Recall: 0.9148
- F1: 0.8878
- Accuracy: 0.9787
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: 30
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 1.0 | 88 | 0.3132 | 0.5506 | 0.4970 | 0.5224 | 0.9194 |
No log | 2.0 | 176 | 0.1509 | 0.7119 | 0.7667 | 0.7383 | 0.9591 |
No log | 3.0 | 264 | 0.1047 | 0.7981 | 0.8499 | 0.8232 | 0.9705 |
No log | 4.0 | 352 | 0.0787 | 0.8314 | 0.8905 | 0.8599 | 0.9770 |
No log | 5.0 | 440 | 0.0802 | 0.9033 | 0.8905 | 0.8968 | 0.9806 |
0.2051 | 6.0 | 528 | 0.0752 | 0.8420 | 0.9189 | 0.8788 | 0.9770 |
0.2051 | 7.0 | 616 | 0.0704 | 0.8854 | 0.9087 | 0.8969 | 0.9823 |
0.2051 | 8.0 | 704 | 0.0732 | 0.8939 | 0.9229 | 0.9082 | 0.9814 |
0.2051 | 9.0 | 792 | 0.0801 | 0.8656 | 0.9148 | 0.8895 | 0.9792 |
0.2051 | 10.0 | 880 | 0.0669 | 0.9192 | 0.9229 | 0.9211 | 0.9850 |
0.2051 | 11.0 | 968 | 0.0783 | 0.8851 | 0.9067 | 0.8958 | 0.9801 |
0.0035 | 12.0 | 1056 | 0.0914 | 0.8542 | 0.9148 | 0.8834 | 0.9780 |
0.0035 | 13.0 | 1144 | 0.1002 | 0.8414 | 0.9148 | 0.8766 | 0.9770 |
0.0035 | 14.0 | 1232 | 0.0978 | 0.8442 | 0.9229 | 0.8818 | 0.9777 |
0.0035 | 15.0 | 1320 | 0.0748 | 0.8830 | 0.9189 | 0.9006 | 0.9816 |
0.0035 | 16.0 | 1408 | 0.0830 | 0.8674 | 0.9026 | 0.8847 | 0.9787 |
0.0035 | 17.0 | 1496 | 0.0938 | 0.8596 | 0.9189 | 0.8882 | 0.9792 |
0.0013 | 18.0 | 1584 | 0.0919 | 0.8651 | 0.9108 | 0.8874 | 0.9792 |
0.0013 | 19.0 | 1672 | 0.0873 | 0.8656 | 0.9148 | 0.8895 | 0.9792 |
0.0013 | 20.0 | 1760 | 0.0888 | 0.8656 | 0.9148 | 0.8895 | 0.9792 |
0.0013 | 21.0 | 1848 | 0.0851 | 0.8685 | 0.9108 | 0.8891 | 0.9792 |
0.0013 | 22.0 | 1936 | 0.0940 | 0.8623 | 0.9148 | 0.8878 | 0.9787 |
0.0005 | 23.0 | 2024 | 0.0845 | 0.8826 | 0.9148 | 0.8984 | 0.9811 |
0.0005 | 24.0 | 2112 | 0.0911 | 0.8690 | 0.9148 | 0.8913 | 0.9792 |
0.0005 | 25.0 | 2200 | 0.0915 | 0.8787 | 0.9108 | 0.8944 | 0.9806 |
0.0005 | 26.0 | 2288 | 0.0951 | 0.8651 | 0.9108 | 0.8874 | 0.9787 |
0.0005 | 27.0 | 2376 | 0.0949 | 0.8585 | 0.9108 | 0.8839 | 0.9782 |
0.0005 | 28.0 | 2464 | 0.0949 | 0.8623 | 0.9148 | 0.8878 | 0.9787 |
0.0005 | 29.0 | 2552 | 0.0946 | 0.8623 | 0.9148 | 0.8878 | 0.9787 |
0.0005 | 30.0 | 2640 | 0.0948 | 0.8623 | 0.9148 | 0.8878 | 0.9787 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu118
- Datasets 2.16.0
- Tokenizers 0.15.0
- Downloads last month
- 5