<!-- 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-aochildes-len-16k-punc-dot
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 3.1868
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 |
|---|---|---|---|
| 6.7347 | 0.29 | 500 | 5.6594 |
| 5.3783 | 0.59 | 1000 | 5.2121 |
| 5.0252 | 0.88 | 1500 | 4.9610 |
| 4.7546 | 1.18 | 2000 | 4.8238 |
| 4.5897 | 1.47 | 2500 | 4.6965 |
| 4.4789 | 1.77 | 3000 | 4.5879 |
| 4.3473 | 2.06 | 3500 | 4.5156 |
| 4.1614 | 2.35 | 4000 | 4.4620 |
| 4.1298 | 2.65 | 4500 | 4.4035 |
| 4.0926 | 2.94 | 5000 | 4.3498 |
| 3.873 | 3.24 | 5500 | 4.3486 |
| 3.8259 | 3.53 | 6000 | 4.3189 |
| 3.809 | 3.83 | 6500 | 4.2819 |
| 3.6844 | 4.12 | 7000 | 4.2885 |
| 3.5391 | 4.41 | 7500 | 4.2779 |
| 3.5315 | 4.71 | 8000 | 4.2655 |
| 3.5178 | 5.0 | 8500 | 4.2534 |
| 3.3396 | 5.3 | 9000 | 4.2694 |
| 3.3435 | 5.59 | 9500 | 4.2672 |
| 3.3344 | 5.89 | 10000 | 4.2660 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
