<!-- 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-cocnat-aochildes-mod-no-repreating-sub-5p9k-length-15p5k
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.1676
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 |
---|---|---|---|
6.7021 | 0.29 | 500 | 5.6361 |
5.3399 | 0.59 | 1000 | 5.1918 |
4.9885 | 0.88 | 1500 | 4.9544 |
4.7127 | 1.18 | 2000 | 4.8024 |
4.5509 | 1.47 | 2500 | 4.6823 |
4.4486 | 1.77 | 3000 | 4.5821 |
4.319 | 2.06 | 3500 | 4.5085 |
4.1344 | 2.36 | 4000 | 4.4554 |
4.0993 | 2.65 | 4500 | 4.4028 |
4.0594 | 2.95 | 5000 | 4.3444 |
3.835 | 3.24 | 5500 | 4.3518 |
3.8041 | 3.54 | 6000 | 4.3167 |
3.7828 | 3.83 | 6500 | 4.2879 |
3.6505 | 4.13 | 7000 | 4.2918 |
3.5174 | 4.42 | 7500 | 4.2869 |
3.5057 | 4.72 | 8000 | 4.2793 |
3.4889 | 5.01 | 8500 | 4.2749 |
3.322 | 5.31 | 9000 | 4.2842 |
3.3162 | 5.6 | 9500 | 4.2824 |
3.3155 | 5.9 | 10000 | 4.2824 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3