metadata
license: mit
base_model: bert-base-german-cased
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: finetuned-bert-base-german-cased
results: []
finetuned-bert-base-german-cased
This model is a fine-tuned version of bert-base-german-cased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4021
- Accuracy: 0.9076
- F1: 0.9079
- Per Class F1: {'Web': 0.973293768545994, 'Panorama': 0.8637681159420288, 'International': 0.9078498293515358, 'Wirtschaft': 0.891304347826087, 'Sport': 0.9916666666666667, 'Inland': 0.825242718446602, 'Etat': 0.9160305343511451, 'Wissenschaft': 0.8717948717948718, 'Kultur': 0.8828828828828829}
- Per Class Accuracy: {'Web': 0.9704142011834319, 'Panorama': 0.8418079096045198, 'International': 0.9366197183098591, 'Wirtschaft': 0.9111111111111111, 'Sport': 0.9916666666666667, 'Inland': 0.8173076923076923, 'Etat': 0.9375, 'Wissenschaft': 0.85, 'Kultur': 0.8596491228070176}
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: 1e-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 | Accuracy | F1 | Per Class F1 | Per Class Accuracy |
|---|---|---|---|---|---|---|---|
| 0.525 | 1.0 | 1156 | 0.3350 | 0.8920 | 0.8919 | {'Web': 0.9583333333333334, 'Panorama': 0.8402366863905326, 'International': 0.8933333333333333, 'Wirtschaft': 0.8741258741258741, 'Sport': 0.9876543209876543, 'Inland': 0.8309178743961353, 'Etat': 0.8617886178861788, 'Wissenschaft': 0.8648648648648649, 'Kultur': 0.8571428571428572} | {'Web': 0.9583333333333334, 'Panorama': 0.8352941176470589, 'International': 0.8993288590604027, 'Wirtschaft': 0.8620689655172413, 'Sport': 0.975609756097561, 'Inland': 0.819047619047619, 'Etat': 0.9464285714285714, 'Wissenschaft': 0.8888888888888888, 'Kultur': 0.8275862068965517} |
| 0.3553 | 2.0 | 2312 | 0.3731 | 0.9086 | 0.9090 | {'Web': 0.960960960960961, 'Panorama': 0.8703170028818443, 'International': 0.9041095890410958, 'Wirtschaft': 0.8970588235294118, 'Sport': 0.995850622406639, 'Inland': 0.8396226415094339, 'Etat': 0.9104477611940298, 'Wissenschaft': 0.8793103448275862, 'Kultur': 0.8807339449541284} | {'Web': 0.9696969696969697, 'Panorama': 0.8435754189944135, 'International': 0.9361702127659575, 'Wirtschaft': 0.9312977099236641, 'Sport': 0.9917355371900827, 'Inland': 0.8090909090909091, 'Etat': 0.9104477611940298, 'Wissenschaft': 0.864406779661017, 'Kultur': 0.8727272727272727} |
| 0.3083 | 3.0 | 3468 | 0.4021 | 0.9076 | 0.9079 | {'Web': 0.973293768545994, 'Panorama': 0.8637681159420288, 'International': 0.9078498293515358, 'Wirtschaft': 0.891304347826087, 'Sport': 0.9916666666666667, 'Inland': 0.825242718446602, 'Etat': 0.9160305343511451, 'Wissenschaft': 0.8717948717948718, 'Kultur': 0.8828828828828829} | {'Web': 0.9704142011834319, 'Panorama': 0.8418079096045198, 'International': 0.9366197183098591, 'Wirtschaft': 0.9111111111111111, 'Sport': 0.9916666666666667, 'Inland': 0.8173076923076923, 'Etat': 0.9375, 'Wissenschaft': 0.85, 'Kultur': 0.8596491228070176} |
Framework versions
- Transformers 4.40.2
- Pytorch 2.3.1
- Datasets 2.19.1
- Tokenizers 0.19.1