<!-- 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. -->
aggregate-all-best-so-far
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 4.3995
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 |
---|---|---|---|
6.686 | 0.3 | 500 | 5.6397 |
5.3431 | 0.6 | 1000 | 5.2192 |
5.0064 | 0.89 | 1500 | 4.9772 |
4.7469 | 1.19 | 2000 | 4.8431 |
4.5938 | 1.49 | 2500 | 4.7258 |
4.4972 | 1.79 | 3000 | 4.6345 |
4.3601 | 2.08 | 3500 | 4.5766 |
4.2 | 2.38 | 4000 | 4.5205 |
4.1717 | 2.68 | 4500 | 4.4612 |
4.1257 | 2.98 | 5000 | 4.4102 |
3.8873 | 3.28 | 5500 | 4.4068 |
3.8774 | 3.57 | 6000 | 4.3738 |
3.8522 | 3.87 | 6500 | 4.3392 |
3.6911 | 4.17 | 7000 | 4.3476 |
3.5905 | 4.47 | 7500 | 4.3367 |
3.5827 | 4.76 | 8000 | 4.3230 |
3.5304 | 5.06 | 8500 | 4.3246 |
3.3915 | 5.36 | 9000 | 4.3290 |
3.4003 | 5.66 | 9500 | 4.3258 |
3.3934 | 5.96 | 10000 | 4.3253 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3