<!-- 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-cbt-rarity-all-no-cbt-7k-p8k
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.1829
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 |
---|---|---|---|
6.7276 | 0.29 | 500 | 5.6334 |
5.3681 | 0.59 | 1000 | 5.2101 |
5.0257 | 0.88 | 1500 | 4.9484 |
4.7492 | 1.17 | 2000 | 4.7957 |
4.586 | 1.47 | 2500 | 4.6672 |
4.476 | 1.76 | 3000 | 4.5561 |
4.3388 | 2.05 | 3500 | 4.4815 |
4.1528 | 2.35 | 4000 | 4.4321 |
4.1203 | 2.64 | 4500 | 4.3692 |
4.08 | 2.93 | 5000 | 4.3225 |
3.8694 | 3.23 | 5500 | 4.3200 |
3.8133 | 3.52 | 6000 | 4.2856 |
3.8082 | 3.81 | 6500 | 4.2543 |
3.6896 | 4.11 | 7000 | 4.2513 |
3.5344 | 4.4 | 7500 | 4.2450 |
3.5282 | 4.69 | 8000 | 4.2309 |
3.5178 | 4.99 | 8500 | 4.2181 |
3.346 | 5.28 | 9000 | 4.2297 |
3.3387 | 5.57 | 9500 | 4.2294 |
3.3317 | 5.87 | 10000 | 4.2282 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3