<!-- 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-cut
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 4.3042
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 |
---|---|---|---|
6.7012 | 0.29 | 500 | 5.6350 |
5.3399 | 0.58 | 1000 | 5.1969 |
4.9839 | 0.87 | 1500 | 4.9406 |
4.7034 | 1.16 | 2000 | 4.7940 |
4.5463 | 1.46 | 2500 | 4.6801 |
4.4423 | 1.75 | 3000 | 4.5644 |
4.3263 | 2.04 | 3500 | 4.4872 |
4.1157 | 2.33 | 4000 | 4.4394 |
4.0929 | 2.62 | 4500 | 4.3840 |
4.0599 | 2.91 | 5000 | 4.3306 |
3.8592 | 3.2 | 5500 | 4.3227 |
3.793 | 3.49 | 6000 | 4.2915 |
3.7801 | 3.79 | 6500 | 4.2609 |
3.6929 | 4.08 | 7000 | 4.2583 |
3.5075 | 4.37 | 7500 | 4.2539 |
3.5083 | 4.66 | 8000 | 4.2380 |
3.4906 | 4.95 | 8500 | 4.2264 |
3.3427 | 5.24 | 9000 | 4.2369 |
3.3099 | 5.53 | 9500 | 4.2359 |
3.3141 | 5.82 | 10000 | 4.2348 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3