<!-- 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-12k-p8k
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.1831
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 |
---|---|---|---|
6.7231 | 0.3 | 500 | 5.6302 |
5.3719 | 0.59 | 1000 | 5.2055 |
5.0197 | 0.89 | 1500 | 4.9471 |
4.7464 | 1.18 | 2000 | 4.7963 |
4.5886 | 1.48 | 2500 | 4.6741 |
4.4756 | 1.78 | 3000 | 4.5652 |
4.3281 | 2.07 | 3500 | 4.4954 |
4.1589 | 2.37 | 4000 | 4.4441 |
4.1235 | 2.66 | 4500 | 4.3829 |
4.0853 | 2.96 | 5000 | 4.3318 |
3.8541 | 3.25 | 5500 | 4.3268 |
3.8262 | 3.55 | 6000 | 4.2947 |
3.8025 | 3.85 | 6500 | 4.2636 |
3.6668 | 4.14 | 7000 | 4.2639 |
3.532 | 4.44 | 7500 | 4.2560 |
3.5262 | 4.73 | 8000 | 4.2409 |
3.4988 | 5.03 | 8500 | 4.2379 |
3.3407 | 5.33 | 9000 | 4.2431 |
3.337 | 5.62 | 9500 | 4.2420 |
3.3339 | 5.92 | 10000 | 4.2411 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3