mongolian-bert-base-multilingual-cased-ner
This model is a fine-tuned version of bert-base-multilingual-cased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1428
- Precision: 0.9085
- Recall: 0.9203
- F1: 0.9143
- Accuracy: 0.9762
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: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.1768 | 1.0 | 477 | 0.0930 | 0.8660 | 0.8939 | 0.8797 | 0.9704 |
| 0.0856 | 2.0 | 954 | 0.0879 | 0.8849 | 0.9082 | 0.8964 | 0.9736 |
| 0.0583 | 3.0 | 1431 | 0.0879 | 0.8905 | 0.9111 | 0.9007 | 0.9749 |
| 0.0404 | 4.0 | 1908 | 0.1053 | 0.8945 | 0.9136 | 0.9040 | 0.9731 |
| 0.0288 | 5.0 | 2385 | 0.1096 | 0.9044 | 0.9144 | 0.9094 | 0.9755 |
| 0.0196 | 6.0 | 2862 | 0.1237 | 0.9045 | 0.9176 | 0.9110 | 0.9754 |
| 0.014 | 7.0 | 3339 | 0.1289 | 0.9066 | 0.9187 | 0.9126 | 0.9757 |
| 0.0099 | 8.0 | 3816 | 0.1342 | 0.9057 | 0.9196 | 0.9126 | 0.9760 |
| 0.0065 | 9.0 | 4293 | 0.1396 | 0.9095 | 0.9195 | 0.9145 | 0.9761 |
| 0.005 | 10.0 | 4770 | 0.1428 | 0.9085 | 0.9203 | 0.9143 | 0.9762 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3