<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
gpt2-dp-cl-rarity
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 4.7719
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: 0.0005
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 1000
- num_epochs: 7
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 6.7749 | 0.27 | 500 | 5.8089 |
| 5.4247 | 0.53 | 1000 | 5.4313 |
| 5.0622 | 0.8 | 1500 | 5.2244 |
| 4.7874 | 1.07 | 2000 | 5.1033 |
| 4.5968 | 1.33 | 2500 | 5.0154 |
| 4.5038 | 1.6 | 3000 | 4.9414 |
| 4.4187 | 1.87 | 3500 | 4.8633 |
| 4.2538 | 2.13 | 4000 | 4.8506 |
| 4.1481 | 2.4 | 4500 | 4.8187 |
| 4.1203 | 2.67 | 5000 | 4.7835 |
| 4.0912 | 2.93 | 5500 | 4.7395 |
| 3.8956 | 3.2 | 6000 | 4.7446 |
| 3.8652 | 3.47 | 6500 | 4.7275 |
| 3.8507 | 3.74 | 7000 | 4.7142 |
| 3.8345 | 4.0 | 7500 | 4.6988 |
| 3.5879 | 4.27 | 8000 | 4.7267 |
| 3.5976 | 4.54 | 8500 | 4.7226 |
| 3.5887 | 4.8 | 9000 | 4.7138 |
| 3.5245 | 5.07 | 9500 | 4.7321 |
| 3.3551 | 5.34 | 10000 | 4.7423 |
| 3.3637 | 5.6 | 10500 | 4.7457 |
| 3.3497 | 5.87 | 11000 | 4.7483 |
| 3.2842 | 6.14 | 11500 | 4.7694 |
| 3.21 | 6.4 | 12000 | 4.7680 |
| 3.2133 | 6.67 | 12500 | 4.7721 |
| 3.2151 | 6.94 | 13000 | 4.7719 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
