<!-- 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. -->
pt2-concat-aochildes-len-16k-rarity-all-6k-1p2k
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.1905
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 |
---|---|---|---|
6.7329 | 0.3 | 500 | 5.6423 |
5.3766 | 0.59 | 1000 | 5.2009 |
5.0318 | 0.89 | 1500 | 4.9558 |
4.7483 | 1.18 | 2000 | 4.8105 |
4.5921 | 1.48 | 2500 | 4.6894 |
4.4881 | 1.77 | 3000 | 4.5781 |
4.3515 | 2.07 | 3500 | 4.5045 |
4.1694 | 2.37 | 4000 | 4.4549 |
4.1375 | 2.66 | 4500 | 4.3951 |
4.0932 | 2.96 | 5000 | 4.3412 |
3.8651 | 3.25 | 5500 | 4.3414 |
3.8328 | 3.55 | 6000 | 4.3062 |
3.8155 | 3.84 | 6500 | 4.2751 |
3.6768 | 4.14 | 7000 | 4.2813 |
3.5406 | 4.44 | 7500 | 4.2701 |
3.533 | 4.73 | 8000 | 4.2561 |
3.5095 | 5.03 | 8500 | 4.2528 |
3.3442 | 5.32 | 9000 | 4.2591 |
3.3407 | 5.62 | 9500 | 4.2577 |
3.344 | 5.91 | 10000 | 4.2571 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3