german-sentiment-bert-finetuned-augmentation-deu-finetuned-augmentation-LUNAR
This model is a fine-tuned version of sercetexam9/german-sentiment-bert-finetuned-augmentation on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3510
- F1: 0.6833
- Roc Auc: 0.7953
- Accuracy: 0.6179
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: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Roc Auc | Accuracy |
|---|---|---|---|---|---|---|
| 0.2285 | 1.0 | 271 | 0.2831 | 0.5569 | 0.7311 | 0.5661 |
| 0.1731 | 2.0 | 542 | 0.2833 | 0.5957 | 0.7509 | 0.5939 |
| 0.1539 | 3.0 | 813 | 0.2965 | 0.6412 | 0.7744 | 0.5772 |
| 0.1303 | 4.0 | 1084 | 0.3158 | 0.6371 | 0.7716 | 0.5689 |
| 0.1144 | 5.0 | 1355 | 0.3013 | 0.6428 | 0.7732 | 0.6050 |
| 0.078 | 6.0 | 1626 | 0.3097 | 0.6405 | 0.7740 | 0.6004 |
| 0.0733 | 7.0 | 1897 | 0.3190 | 0.6445 | 0.7775 | 0.5994 |
| 0.0578 | 8.0 | 2168 | 0.3320 | 0.6550 | 0.7836 | 0.5865 |
| 0.0514 | 9.0 | 2439 | 0.3380 | 0.6557 | 0.7803 | 0.5957 |
| 0.0418 | 10.0 | 2710 | 0.3350 | 0.6723 | 0.7905 | 0.6068 |
| 0.0445 | 11.0 | 2981 | 0.3451 | 0.6631 | 0.7863 | 0.6087 |
| 0.028 | 12.0 | 3252 | 0.3464 | 0.6674 | 0.7852 | 0.6179 |
| 0.0256 | 13.0 | 3523 | 0.3510 | 0.6833 | 0.7953 | 0.6179 |
| 0.0197 | 14.0 | 3794 | 0.3604 | 0.6663 | 0.7844 | 0.6096 |
| 0.0169 | 15.0 | 4065 | 0.3651 | 0.6783 | 0.7950 | 0.6235 |
| 0.0119 | 16.0 | 4336 | 0.3744 | 0.6590 | 0.7816 | 0.6050 |
| 0.0213 | 17.0 | 4607 | 0.3710 | 0.6653 | 0.7847 | 0.6170 |
Framework versions
- Transformers 4.45.1
- Pytorch 2.4.0
- Datasets 3.0.1
- Tokenizers 0.20.0