<!-- 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-aochildes-len-16k-rarity-all-3k-p95k
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.1918
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 |
|---|---|---|---|
| 6.7409 | 0.29 | 500 | 5.6340 |
| 5.3726 | 0.59 | 1000 | 5.2051 |
| 5.0259 | 0.88 | 1500 | 4.9531 |
| 4.7454 | 1.18 | 2000 | 4.8054 |
| 4.5934 | 1.47 | 2500 | 4.6754 |
| 4.4864 | 1.77 | 3000 | 4.5725 |
| 4.3481 | 2.06 | 3500 | 4.5028 |
| 4.1699 | 2.36 | 4000 | 4.4581 |
| 4.1348 | 2.65 | 4500 | 4.3968 |
| 4.0906 | 2.95 | 5000 | 4.3403 |
| 3.8679 | 3.24 | 5500 | 4.3395 |
| 3.8308 | 3.54 | 6000 | 4.3080 |
| 3.8137 | 3.83 | 6500 | 4.2756 |
| 3.6811 | 4.13 | 7000 | 4.2786 |
| 3.5439 | 4.42 | 7500 | 4.2680 |
| 3.5384 | 4.72 | 8000 | 4.2581 |
| 3.5122 | 5.01 | 8500 | 4.2522 |
| 3.3498 | 5.31 | 9000 | 4.2589 |
| 3.3434 | 5.6 | 9500 | 4.2583 |
| 3.3411 | 5.9 | 10000 | 4.2578 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
