DOOGLAK's picture
update model card README.md
98d7003
metadata
license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - article100v4_wikigold_split
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: Article_100v4_NER_Model_3Epochs_UNAUGMENTED
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: article100v4_wikigold_split
          type: article100v4_wikigold_split
          args: default
        metrics:
          - name: Precision
            type: precision
            value: 0.16216216216216217
          - name: Recall
            type: recall
            value: 0.0030761343245321714
          - name: F1
            type: f1
            value: 0.006037735849056605
          - name: Accuracy
            type: accuracy
            value: 0.7826991539137269

Article_100v4_NER_Model_3Epochs_UNAUGMENTED

This model is a fine-tuned version of bert-base-cased on the article100v4_wikigold_split dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5862
  • Precision: 0.1622
  • Recall: 0.0031
  • F1: 0.0060
  • Accuracy: 0.7827

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 3

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 1.0 12 0.7284 0.0 0.0 0.0 0.7816
No log 2.0 24 0.6227 0.2 0.0003 0.0005 0.7817
No log 3.0 36 0.5862 0.1622 0.0031 0.0060 0.7827

Framework versions

  • Transformers 4.17.0
  • Pytorch 1.11.0+cu113
  • Datasets 2.4.0
  • Tokenizers 0.11.6