<!-- 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-bnc-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: 4.3248
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 |
|---|---|---|---|
| 6.7001 | 0.29 | 500 | 5.6381 |
| 5.3463 | 0.58 | 1000 | 5.2011 |
| 5.0002 | 0.88 | 1500 | 4.9537 |
| 4.7179 | 1.17 | 2000 | 4.8088 |
| 4.5667 | 1.46 | 2500 | 4.6879 |
| 4.4591 | 1.75 | 3000 | 4.5783 |
| 4.3421 | 2.05 | 3500 | 4.5052 |
| 4.1384 | 2.34 | 4000 | 4.4566 |
| 4.1141 | 2.63 | 4500 | 4.4035 |
| 4.0722 | 2.92 | 5000 | 4.3454 |
| 3.872 | 3.22 | 5500 | 4.3418 |
| 3.8179 | 3.51 | 6000 | 4.3082 |
| 3.7927 | 3.8 | 6500 | 4.2809 |
| 3.691 | 4.09 | 7000 | 4.2801 |
| 3.527 | 4.39 | 7500 | 4.2714 |
| 3.5142 | 4.68 | 8000 | 4.2548 |
| 3.5077 | 4.97 | 8500 | 4.2408 |
| 3.3436 | 5.26 | 9000 | 4.2581 |
| 3.3327 | 5.56 | 9500 | 4.2562 |
| 3.3282 | 5.85 | 10000 | 4.2551 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
