<!-- 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-concat-wiki-rarity-no-cut
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 4.3201
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: 6
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 6.7051 | 0.29 | 500 | 5.6378 |
| 5.3367 | 0.58 | 1000 | 5.1972 |
| 4.9867 | 0.87 | 1500 | 4.9538 |
| 4.7104 | 1.16 | 2000 | 4.8093 |
| 4.5621 | 1.46 | 2500 | 4.6885 |
| 4.4544 | 1.75 | 3000 | 4.5808 |
| 4.3353 | 2.04 | 3500 | 4.5031 |
| 4.1291 | 2.33 | 4000 | 4.4542 |
| 4.1138 | 2.62 | 4500 | 4.3959 |
| 4.0612 | 2.91 | 5000 | 4.3429 |
| 3.8709 | 3.2 | 5500 | 4.3403 |
| 3.8046 | 3.49 | 6000 | 4.3115 |
| 3.7892 | 3.78 | 6500 | 4.2732 |
| 3.7056 | 4.07 | 7000 | 4.2679 |
| 3.5187 | 4.37 | 7500 | 4.2666 |
| 3.5135 | 4.66 | 8000 | 4.2503 |
| 3.5039 | 4.95 | 8500 | 4.2386 |
| 3.3508 | 5.24 | 9000 | 4.2509 |
| 3.324 | 5.53 | 9500 | 4.2505 |
| 3.3217 | 5.82 | 10000 | 4.2496 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
