<!-- 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-mod-datasets1-rarity-all-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: 4.8800
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 |
|---|---|---|---|
| 6.8093 | 0.32 | 500 | 5.7720 |
| 5.5016 | 0.63 | 1000 | 5.4100 |
| 5.1549 | 0.95 | 1500 | 5.2064 |
| 4.875 | 1.26 | 2000 | 5.0988 |
| 4.7498 | 1.58 | 2500 | 5.0408 |
| 4.6514 | 1.89 | 3000 | 4.9468 |
| 4.4676 | 2.21 | 3500 | 4.9295 |
| 4.3873 | 2.52 | 4000 | 4.8977 |
| 4.339 | 2.84 | 4500 | 4.8540 |
| 4.1886 | 3.15 | 5000 | 4.8392 |
| 4.0719 | 3.47 | 5500 | 4.8195 |
| 4.0517 | 3.78 | 6000 | 4.8159 |
| 3.9509 | 4.1 | 6500 | 4.8221 |
| 3.7784 | 4.41 | 7000 | 4.8293 |
| 3.7677 | 4.73 | 7500 | 4.8194 |
| 3.7313 | 5.04 | 8000 | 4.8217 |
| 3.5746 | 5.36 | 8500 | 4.8334 |
| 3.569 | 5.67 | 9000 | 4.8353 |
| 3.5733 | 5.99 | 9500 | 4.8343 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
