BahAdoR0101's picture
End of training
464161d
metadata
license: mit
base_model: dslim/bert-base-NER
tags:
  - generated_from_trainer
datasets:
  - conll2003job
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: my_xlm-roberta-large-finetuned-conlljob01
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: conll2003job
          type: conll2003job
          config: conll2003job
          split: test
          args: conll2003job
        metrics:
          - name: Precision
            type: precision
            value: 0.9057427125152732
          - name: Recall
            type: recall
            value: 0.9187322946175638
          - name: F1
            type: f1
            value: 0.9121912630746243
          - name: Accuracy
            type: accuracy
            value: 0.9825347259610208

my_xlm-roberta-large-finetuned-conlljob01

This model is a fine-tuned version of dslim/bert-base-NER on the conll2003job dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1690
  • Precision: 0.9057
  • Recall: 0.9187
  • F1: 0.9122
  • Accuracy: 0.9825

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
0.0372 1.0 896 0.1439 0.8943 0.9184 0.9062 0.9816
0.0043 2.0 1792 0.1532 0.9047 0.9209 0.9127 0.9824
0.0019 3.0 2688 0.1652 0.9102 0.9186 0.9143 0.9828
0.0013 4.0 3584 0.1690 0.9057 0.9187 0.9122 0.9825

Framework versions

  • Transformers 4.34.1
  • Pytorch 2.1.0+cu118
  • Datasets 2.14.5
  • Tokenizers 0.14.1