gpt_16_5_5.6e-5_lp5_nb10
This model is a fine-tuned version of skt/kogpt2-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 4.0126
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: 5.6e-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: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 5.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 3.1407 | 0.38 | 1000 | 4.0864 |
| 3.1639 | 0.76 | 2000 | 4.0867 |
| 2.87 | 1.13 | 3000 | 4.0611 |
| 2.8591 | 1.51 | 4000 | 4.0315 |
| 2.8866 | 1.89 | 5000 | 4.0079 |
| 2.6146 | 2.27 | 6000 | 4.0153 |
| 2.6576 | 2.64 | 7000 | 4.0021 |
| 2.5361 | 3.02 | 8000 | 4.0106 |
| 2.4552 | 3.4 | 9000 | 4.0137 |
| 2.4595 | 3.78 | 10000 | 3.9967 |
| 2.3286 | 4.15 | 11000 | 4.0204 |
| 2.2805 | 4.53 | 12000 | 4.0154 |
| 2.3098 | 4.91 | 13000 | 4.0126 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.9.0+cu102
- Datasets 2.8.0
- Tokenizers 0.13.2