<!-- 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_simple_wikipedia
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.9720
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: 128
- eval_batch_size: 128
- 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 |
|---|---|---|---|
| 5.7673 | 0.29 | 500 | 4.8493 |
| 4.5669 | 0.57 | 1000 | 4.5137 |
| 4.3049 | 0.86 | 1500 | 4.3240 |
| 4.0824 | 1.14 | 2000 | 4.2154 |
| 3.9603 | 1.43 | 2500 | 4.1236 |
| 3.8809 | 1.71 | 3000 | 4.0511 |
| 3.8162 | 2.0 | 3500 | 3.9894 |
| 3.6371 | 2.28 | 4000 | 3.9735 |
| 3.6156 | 2.57 | 4500 | 3.9275 |
| 3.5854 | 2.85 | 5000 | 3.8877 |
| 3.4747 | 3.14 | 5500 | 3.8865 |
| 3.3896 | 3.42 | 6000 | 3.8698 |
| 3.3947 | 3.71 | 6500 | 3.8381 |
| 3.385 | 4.0 | 7000 | 3.8069 |
| 3.1618 | 4.28 | 7500 | 3.8443 |
| 3.1943 | 4.57 | 8000 | 3.8247 |
| 3.1957 | 4.85 | 8500 | 3.8051 |
| 3.0617 | 5.14 | 9000 | 3.8418 |
| 2.9657 | 5.42 | 9500 | 3.8426 |
| 2.9933 | 5.71 | 10000 | 3.8303 |
| 2.9946 | 5.99 | 10500 | 3.8144 |
| 2.7379 | 6.28 | 11000 | 3.8735 |
| 2.7665 | 6.56 | 11500 | 3.8718 |
| 2.779 | 6.85 | 12000 | 3.8670 |
| 2.6666 | 7.13 | 12500 | 3.9040 |
| 2.5564 | 7.42 | 13000 | 3.9169 |
| 2.5723 | 7.71 | 13500 | 3.9197 |
| 2.5749 | 7.99 | 14000 | 3.9177 |
| 2.4112 | 8.28 | 14500 | 3.9522 |
| 2.4142 | 8.56 | 15000 | 3.9562 |
| 2.4162 | 8.85 | 15500 | 3.9600 |
| 2.3796 | 9.13 | 16000 | 3.9681 |
| 2.3344 | 9.42 | 16500 | 3.9714 |
| 2.3397 | 9.7 | 17000 | 3.9721 |
| 2.3387 | 9.99 | 17500 | 3.9720 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
