<!-- 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. -->
Article_100v1_NER_Model_3Epochs_AUGMENTED
This model is a fine-tuned version of bert-base-cased on the article100v1_wikigold_split dataset. It achieves the following results on the evaluation set:
- Loss: 0.3140
- Precision: 0.4708
- Recall: 0.4550
- F1: 0.4628
- Accuracy: 0.8932
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 | 40 | 0.4092 | 0.1933 | 0.1445 | 0.1654 | 0.8398 |
| No log | 2.0 | 80 | 0.3279 | 0.4254 | 0.3924 | 0.4082 | 0.8851 |
| No log | 3.0 | 120 | 0.3140 | 0.4708 | 0.4550 | 0.4628 | 0.8932 |
Framework versions
- Transformers 4.17.0
- Pytorch 1.11.0+cu113
- Datasets 2.4.0
- Tokenizers 0.11.6
