en-multinerd-ner-roberta
This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0280
- Precision: 0.9421
- Recall: 0.9446
- F1: 0.9434
- Accuracy: 0.9903
- Per-precision: 0.9917
- Per-recall: 0.9970
- Per-f1: 0.9943
- Org-precision: 0.9766
- Org-recall: 0.9837
- Org-f1: 0.9801
- Loc-precision: 0.9959
- Loc-recall: 0.9934
- Loc-f1: 0.9947
- Dis-precision: 0.7665
- Dis-recall: 0.7655
- Dis-f1: 0.7660
- Anim-precision: 0.6897
- Anim-recall: 0.7039
- Anim-f1: 0.6967
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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | Per-precision | Per-recall | Per-f1 | Org-precision | Org-recall | Org-f1 | Loc-precision | Loc-recall | Loc-f1 | Dis-precision | Dis-recall | Dis-f1 | Anim-precision | Anim-recall | Anim-f1 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
0.038 | 1.0 | 8205 | 0.0452 | 0.8742 | 0.9240 | 0.8984 | 0.9830 | 0.9963 | 0.9955 | 0.9959 | 0.9616 | 0.9795 | 0.9705 | 0.9894 | 0.9944 | 0.9919 | 0.6658 | 0.7487 | 0.7048 | 0.6415 | 0.7857 | 0.7063 |
0.0294 | 2.0 | 16410 | 0.0410 | 0.9154 | 0.9085 | 0.9119 | 0.9856 | 0.9952 | 0.9980 | 0.9966 | 0.9697 | 0.9814 | 0.9755 | 0.9929 | 0.9946 | 0.9938 | 0.7358 | 0.7162 | 0.7259 | 0.7107 | 0.6455 | 0.6765 |
0.0202 | 3.0 | 24615 | 0.0429 | 0.9023 | 0.9255 | 0.9137 | 0.9854 | 0.9950 | 0.9973 | 0.9961 | 0.9791 | 0.9766 | 0.9779 | 0.9905 | 0.9958 | 0.9931 | 0.7190 | 0.7622 | 0.7400 | 0.7247 | 0.7522 | 0.7382 |
0.0149 | 4.0 | 32820 | 0.0455 | 0.9122 | 0.9205 | 0.9163 | 0.9858 | 0.9955 | 0.9971 | 0.9963 | 0.9795 | 0.9781 | 0.9788 | 0.9928 | 0.9954 | 0.9941 | 0.7295 | 0.7605 | 0.7447 | 0.7174 | 0.7363 | 0.7267 |
Framework versions
- Transformers 4.36.1
- Pytorch 2.1.0+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0
- Downloads last month
- 3