<!-- 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_left_out_wikipedia
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.8366
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 |
|---|---|---|---|
| 5.8141 | 0.27 | 500 | 4.8520 |
| 4.5861 | 0.53 | 1000 | 4.4909 |
| 4.3045 | 0.8 | 1500 | 4.2742 |
| 4.0861 | 1.07 | 2000 | 4.1490 |
| 3.9278 | 1.33 | 2500 | 4.0562 |
| 3.8591 | 1.6 | 3000 | 3.9800 |
| 3.7835 | 1.87 | 3500 | 3.9083 |
| 3.6499 | 2.13 | 4000 | 3.8799 |
| 3.567 | 2.4 | 4500 | 3.8381 |
| 3.5361 | 2.67 | 5000 | 3.7975 |
| 3.5278 | 2.93 | 5500 | 3.7552 |
| 3.3555 | 3.2 | 6000 | 3.7622 |
| 3.3265 | 3.47 | 6500 | 3.7426 |
| 3.3305 | 3.73 | 7000 | 3.7122 |
| 3.3246 | 4.0 | 7500 | 3.6889 |
| 3.0968 | 4.27 | 8000 | 3.7216 |
| 3.1248 | 4.53 | 8500 | 3.7057 |
| 3.1354 | 4.8 | 9000 | 3.6846 |
| 3.0701 | 5.07 | 9500 | 3.7066 |
| 2.8974 | 5.33 | 10000 | 3.7183 |
| 2.9258 | 5.6 | 10500 | 3.7096 |
| 2.9387 | 5.87 | 11000 | 3.6943 |
| 2.7975 | 6.13 | 11500 | 3.7369 |
| 2.6972 | 6.4 | 12000 | 3.7468 |
| 2.7193 | 6.67 | 12500 | 3.7422 |
| 2.7233 | 6.93 | 13000 | 3.7337 |
| 2.5434 | 7.2 | 13500 | 3.7783 |
| 2.5072 | 7.47 | 14000 | 3.7864 |
| 2.5183 | 7.73 | 14500 | 3.7869 |
| 2.5263 | 8.0 | 15000 | 3.7838 |
| 2.3533 | 8.27 | 15500 | 3.8174 |
| 2.3661 | 8.53 | 16000 | 3.8220 |
| 2.3659 | 8.8 | 16500 | 3.8246 |
| 2.3462 | 9.07 | 17000 | 3.8313 |
| 2.286 | 9.33 | 17500 | 3.8359 |
| 2.2867 | 9.6 | 18000 | 3.8367 |
| 2.2885 | 9.87 | 18500 | 3.8366 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
