HuggingAlex1247's picture
Upload README.md
a909d42
metadata
language:
  - de
license: mit
datasets:
  - germeval_14
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: distilbert-base-german-europeana-cased-germeval_14
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: germeval_14
          type: germeval_14
          args: default
        metrics:
          - name: precision
            type: precision
            value: 0.7437097717963721
          - name: recall
            type: recall
            value: 0.7571485305798252
          - name: f1
            type: f1
            value: 0.7503689855357669
          - name: accuracy
            type: accuracy
            value: 0.9541357066185896

distilbert-base-german-europeana-cased-germeval_14

This model is a fine-tuned version of dbmdz/distilbert-base-german-europeana-cased on the germeval_14 dataset. It achieves the following results on the evaluation set:

  • precision: 0.7437
  • recall: 0.7571
  • f1: 0.7504
  • accuracy: 0.9541

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:

  • num_train_epochs: 10
  • train_batch_size: 8
  • eval_batch_size: 8
  • learning_rate: 1e-05
  • weight_decay_rate: 0.01
  • num_warmup_steps: 0
  • fp16: True

Framework versions

  • Transformers 4.17.0
  • Datasets 1.18.0
  • Tokenizers 0.11.6