<!-- 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. -->
gpt2-2_left_out_aochildes
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.9760
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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 6.0172 | 0.25 | 500 | 5.1073 |
| 4.785 | 0.5 | 1000 | 4.7224 |
| 4.4827 | 0.75 | 1500 | 4.4942 |
| 4.298 | 0.99 | 2000 | 4.3427 |
| 4.0754 | 1.24 | 2500 | 4.2557 |
| 4.0142 | 1.49 | 3000 | 4.1790 |
| 3.9352 | 1.74 | 3500 | 4.1043 |
| 3.8786 | 1.99 | 4000 | 4.0364 |
| 3.6731 | 2.24 | 4500 | 4.0186 |
| 3.6688 | 2.49 | 5000 | 3.9721 |
| 3.661 | 2.73 | 5500 | 3.9350 |
| 3.6258 | 2.98 | 6000 | 3.8969 |
| 3.4249 | 3.23 | 6500 | 3.9079 |
| 3.425 | 3.48 | 7000 | 3.8847 |
| 3.4348 | 3.73 | 7500 | 3.8593 |
| 3.4292 | 3.98 | 8000 | 3.8315 |
| 3.2039 | 4.23 | 8500 | 3.8684 |
| 3.2172 | 4.48 | 9000 | 3.8502 |
| 3.2328 | 4.72 | 9500 | 3.8328 |
| 3.2348 | 4.97 | 10000 | 3.8186 |
| 2.9915 | 5.22 | 10500 | 3.8657 |
| 3.0125 | 5.47 | 11000 | 3.8594 |
| 3.0221 | 5.72 | 11500 | 3.8446 |
| 3.0287 | 5.97 | 12000 | 3.8301 |
| 2.797 | 6.22 | 12500 | 3.8864 |
| 2.7847 | 6.46 | 13000 | 3.8881 |
| 2.8069 | 6.71 | 13500 | 3.8780 |
| 2.8152 | 6.96 | 14000 | 3.8733 |
| 2.6075 | 7.21 | 14500 | 3.9199 |
| 2.587 | 7.46 | 15000 | 3.9290 |
| 2.6053 | 7.71 | 15500 | 3.9274 |
| 2.6083 | 7.96 | 16000 | 3.9261 |
| 2.4643 | 8.2 | 16500 | 3.9556 |
| 2.4402 | 8.45 | 17000 | 3.9621 |
| 2.4498 | 8.7 | 17500 | 3.9633 |
| 2.4516 | 8.95 | 18000 | 3.9635 |
| 2.3806 | 9.2 | 18500 | 3.9736 |
| 2.3719 | 9.45 | 19000 | 3.9753 |
| 2.372 | 9.7 | 19500 | 3.9759 |
| 2.3701 | 9.95 | 20000 | 3.9760 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
