metadata

library_name: transformers
license: mit
base_model: bert-base-german-cased
tags:
  - generated_from_trainer
datasets:
  - germeval_14
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: ner_model
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: germeval_14
          type: germeval_14
          config: germeval_14
          split: test
          args: germeval_14
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9703112769708805
          - name: F1
            type: f1
            value: 0.8348410033576931
          - name: Precision
            type: precision
            value: 0.8311310366525091
          - name: Recall
            type: recall
            value: 0.8385842393460836

ner_model

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

Loss: 0.1446
Accuracy: 0.9703
F1: 0.8348
Precision: 0.8311
Recall: 0.8386

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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.1274	1.0	3000	0.1132	0.9671	0.8144	0.8031	0.8260
0.065	2.0	6000	0.1382	0.9690	0.8301	0.8452	0.8155
0.0365	3.0	9000	0.1446	0.9703	0.8348	0.8311	0.8386

Framework versions

Transformers 4.47.1
Pytorch 2.5.1+cu124
Datasets 3.2.0
Tokenizers 0.21.0