gpt2Kaggle_v2
This model is a fine-tuned version of ytu-ce-cosmos/turkish-gpt2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 6.6387
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: 0.0001
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 128
- total_train_batch_size: 2048
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
9.3082 | 0.99 | 39 | 8.3464 |
8.005 | 1.98 | 78 | 7.8311 |
7.3738 | 2.99 | 118 | 7.4780 |
7.21 | 3.98 | 157 | 7.1821 |
6.7504 | 4.99 | 197 | 6.9618 |
6.9037 | 5.98 | 236 | 6.8182 |
6.5975 | 6.99 | 276 | 6.7208 |
6.5246 | 7.98 | 315 | 6.6663 |
6.3202 | 8.99 | 355 | 6.6420 |
6.3742 | 9.88 | 390 | 6.6387 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 2