leonadase's picture
update model card README.md
55a9839
metadata
tags:
  - generated_from_trainer
datasets:
  - fdner
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: bert-base-chinese-finetuned-ner-v1
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: fdner
          type: fdner
          args: fdner
        metrics:
          - name: Precision
            type: precision
            value: 0.981203007518797
          - name: Recall
            type: recall
            value: 0.9886363636363636
          - name: F1
            type: f1
            value: 0.9849056603773584
          - name: Accuracy
            type: accuracy
            value: 0.9909536373916321

bert-base-chinese-finetuned-ner-v1

This model is a fine-tuned version of bert-base-chinese on the fdner dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0413
  • Precision: 0.9812
  • Recall: 0.9886
  • F1: 0.9849
  • Accuracy: 0.9910

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: 10
  • eval_batch_size: 10
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 1.0 8 2.0640 0.0 0.0 0.0 0.4323
No log 2.0 16 1.7416 0.0204 0.0227 0.0215 0.5123
No log 3.0 24 1.5228 0.0306 0.0265 0.0284 0.5456
No log 4.0 32 1.2597 0.0961 0.1591 0.1198 0.6491
No log 5.0 40 1.0273 0.1588 0.2159 0.1830 0.7450
No log 6.0 48 0.8026 0.2713 0.3258 0.2960 0.8208
No log 7.0 56 0.6547 0.36 0.4091 0.3830 0.8513
No log 8.0 64 0.5180 0.4650 0.5038 0.4836 0.8873
No log 9.0 72 0.4318 0.5139 0.5606 0.5362 0.9067
No log 10.0 80 0.3511 0.6169 0.6894 0.6512 0.9291
No log 11.0 88 0.2887 0.6691 0.6894 0.6791 0.9414
No log 12.0 96 0.2396 0.7042 0.7576 0.7299 0.9516
No log 13.0 104 0.2052 0.7568 0.8371 0.7950 0.9587
No log 14.0 112 0.1751 0.8303 0.8712 0.8503 0.9610
No log 15.0 120 0.1512 0.8464 0.8977 0.8713 0.9668
No log 16.0 128 0.1338 0.8759 0.9091 0.8922 0.9710
No log 17.0 136 0.1147 0.8959 0.9129 0.9043 0.9746
No log 18.0 144 0.1011 0.9326 0.9432 0.9379 0.9761
No log 19.0 152 0.0902 0.9251 0.9356 0.9303 0.9795
No log 20.0 160 0.0806 0.9440 0.9583 0.9511 0.9804
No log 21.0 168 0.0743 0.9586 0.9659 0.9623 0.9812
No log 22.0 176 0.0649 0.9511 0.9583 0.9547 0.9851
No log 23.0 184 0.0595 0.9591 0.9773 0.9681 0.9876
No log 24.0 192 0.0537 0.9625 0.9735 0.9680 0.9883
No log 25.0 200 0.0505 0.9701 0.9848 0.9774 0.9894
No log 26.0 208 0.0464 0.9737 0.9811 0.9774 0.9904
No log 27.0 216 0.0439 0.9737 0.9811 0.9774 0.9906
No log 28.0 224 0.0428 0.9812 0.9886 0.9849 0.9910
No log 29.0 232 0.0417 0.9812 0.9886 0.9849 0.9910
No log 30.0 240 0.0413 0.9812 0.9886 0.9849 0.9910

Framework versions

  • Transformers 4.18.0
  • Pytorch 1.10.0+cu111
  • Datasets 2.0.0
  • Tokenizers 0.11.6