<!-- 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-cl-rarity-sampling-3
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 4.8082
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: 1
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 6.5861 | 0.04 | 500 | 5.8694 |
| 5.354 | 0.08 | 1000 | 5.4475 |
| 5.023 | 0.11 | 1500 | 5.2264 |
| 4.8006 | 0.15 | 2000 | 5.0886 |
| 4.6394 | 0.19 | 2500 | 5.0063 |
| 4.5152 | 0.23 | 3000 | 4.9501 |
| 4.4117 | 0.27 | 3500 | 4.8973 |
| 4.3195 | 0.3 | 4000 | 4.8588 |
| 4.2286 | 0.34 | 4500 | 4.8358 |
| 4.1463 | 0.38 | 5000 | 4.8088 |
| 4.0689 | 0.42 | 5500 | 4.7887 |
| 3.9901 | 0.46 | 6000 | 4.7805 |
| 3.917 | 0.49 | 6500 | 4.7758 |
| 3.8461 | 0.53 | 7000 | 4.7615 |
| 3.7665 | 0.57 | 7500 | 4.7577 |
| 3.7044 | 0.61 | 8000 | 4.7552 |
| 3.637 | 0.65 | 8500 | 4.7574 |
| 3.573 | 0.68 | 9000 | 4.7594 |
| 3.5162 | 0.72 | 9500 | 4.7603 |
| 3.4583 | 0.76 | 10000 | 4.7634 |
| 3.4217 | 0.8 | 10500 | 4.7641 |
| 3.3828 | 0.83 | 11000 | 4.7636 |
| 3.3569 | 0.87 | 11500 | 4.7628 |
| 3.3358 | 0.91 | 12000 | 4.7636 |
| 3.3235 | 0.95 | 12500 | 4.7638 |
| 3.3223 | 0.99 | 13000 | 4.7636 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
