<!-- 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. -->
finetune_ner
This model is a fine-tuned version of deprem-ml/deprem-ner on a custom private dataset. It achieves the following results on the evaluation set:
- Loss: 0.2600
- Precision: 0.6071
- Recall: 0.68
- F1: 0.6415
- Accuracy: 0.9055
Training procedure
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 1.0 | 2 | 0.3447 | 0.4828 | 0.56 | 0.5185 | 0.8740 |
No log | 2.0 | 4 | 0.3079 | 0.4688 | 0.6 | 0.5263 | 0.9055 |
No log | 3.0 | 6 | 0.2849 | 0.5312 | 0.68 | 0.5965 | 0.9055 |
No log | 4.0 | 8 | 0.2796 | 0.5484 | 0.68 | 0.6071 | 0.8976 |
No log | 5.0 | 10 | 0.2741 | 0.6071 | 0.68 | 0.6415 | 0.9055 |
No log | 6.0 | 12 | 0.2705 | 0.6071 | 0.68 | 0.6415 | 0.9055 |
No log | 7.0 | 14 | 0.2685 | 0.5862 | 0.68 | 0.6296 | 0.9055 |
No log | 8.0 | 16 | 0.2636 | 0.6071 | 0.68 | 0.6415 | 0.9055 |
No log | 9.0 | 18 | 0.2611 | 0.6071 | 0.68 | 0.6415 | 0.9055 |
No log | 10.0 | 20 | 0.2600 | 0.6071 | 0.68 | 0.6415 | 0.9055 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3