<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
bert-finetuned-ner_swedish_test_NUMb_2
This model is a fine-tuned version of KBLab/bert-base-swedish-cased-ner on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0676
- Precision: 0.75
- Recall: 0.7179
- F1: 0.7336
- Accuracy: 0.9811
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: 3
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 1.0 | 128 | 0.0637 | 0.7477 | 0.6838 | 0.7143 | 0.9816 |
No log | 2.0 | 256 | 0.0642 | 0.7304 | 0.7179 | 0.7241 | 0.9803 |
No log | 3.0 | 384 | 0.0676 | 0.75 | 0.7179 | 0.7336 | 0.9811 |
Framework versions
- Transformers 4.19.3
- Pytorch 1.7.1
- Datasets 2.2.2
- Tokenizers 0.12.1