BahAdoR0101's picture
End of training
1087ceb
metadata
license: mit
base_model: dslim/bert-large-NER
tags:
  - generated_from_trainer
datasets:
  - job-titles
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: my_awesome_ner_model
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: job-titles
          type: job-titles
          config: job-titles
          split: test
          args: job-titles
        metrics:
          - name: Precision
            type: precision
            value: 0.9863945578231292
          - name: Recall
            type: recall
            value: 0.9954233409610984
          - name: F1
            type: f1
            value: 0.9908883826879271
          - name: Accuracy
            type: accuracy
            value: 0.9953216374269006

my_awesome_ner_model

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

  • Loss: 0.0080
  • Precision: 0.9864
  • Recall: 0.9954
  • F1: 0.9909
  • Accuracy: 0.9953

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: 2

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 1.0 18 0.0232 0.9864 0.9954 0.9909 0.9953
No log 2.0 36 0.0080 0.9864 0.9954 0.9909 0.9953

Framework versions

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