turkish-gpt2-mini-M2-cleaned-politics720k-10ep
This model is a fine-tuned version of eminAydin/turkish-gpt2-mini-M0-cleaned-wiki720-10ep on the None dataset. It achieves the following results on the evaluation set:
- Loss: 4.1690
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 4.8453 | 1.0 | 9070 | 4.6594 |
| 4.3703 | 2.0 | 18141 | 4.4607 |
| 4.186 | 3.0 | 27212 | 4.3564 |
| 4.0629 | 4.0 | 36283 | 4.2902 |
| 3.9703 | 5.0 | 45354 | 4.2425 |
| 3.8977 | 6.0 | 54425 | 4.2108 |
| 3.8423 | 7.0 | 63496 | 4.1870 |
| 3.802 | 8.0 | 72567 | 4.1752 |
| 3.7763 | 9.0 | 81637 | 4.1695 |
| 3.7628 | 10.0 | 90700 | 4.1690 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2