<!-- 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-mod-sub-1k-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: 4.3376
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 |
|---|---|---|---|
| 6.7163 | 0.29 | 500 | 5.6360 |
| 5.3451 | 0.59 | 1000 | 5.2045 |
| 4.9911 | 0.88 | 1500 | 4.9520 |
| 4.7131 | 1.17 | 2000 | 4.8067 |
| 4.5566 | 1.47 | 2500 | 4.6857 |
| 4.459 | 1.76 | 3000 | 4.5793 |
| 4.3204 | 2.05 | 3500 | 4.5032 |
| 4.1317 | 2.35 | 4000 | 4.4683 |
| 4.1054 | 2.64 | 4500 | 4.4074 |
| 4.0664 | 2.93 | 5000 | 4.3520 |
| 3.8537 | 3.23 | 5500 | 4.3521 |
| 3.8012 | 3.52 | 6000 | 4.3248 |
| 3.7864 | 3.81 | 6500 | 4.2888 |
| 3.6745 | 4.11 | 7000 | 4.2966 |
| 3.5228 | 4.4 | 7500 | 4.2867 |
| 3.5101 | 4.69 | 8000 | 4.2753 |
| 3.5031 | 4.99 | 8500 | 4.2624 |
| 3.3329 | 5.28 | 9000 | 4.2786 |
| 3.325 | 5.58 | 9500 | 4.2790 |
| 3.3197 | 5.87 | 10000 | 4.2789 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
