umit_42000news
Dataset:
https://www.kaggle.com/datasets/furkanozbay/turkish-news-dataset
https://www.kaggle.com/datasets/oktayozturk010/42000-news-text-in-13-classes
This model is a fine-tuned version of dbmdz/bert-base-turkish-cased on provided dataset. It achieves the following results on the evaluation set:
- Loss: 0.9423
- Accuracy: 0.6937
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: 5e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.8754 | 1.0 | 1584 | 0.9817 | 0.6752 |
| 0.7769 | 2.0 | 3168 | 0.9106 | 0.6903 |
| 0.527 | 3.0 | 4752 | 0.9423 | 0.6937 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu117
- Datasets 2.10.1
- Tokenizers 0.13.2
- Downloads last month
- 2