| --- |
| license: mit |
| tags: |
| - generated_from_trainer |
| datasets: |
| - generator |
| model-index: |
| - name: all-base |
| results: [] |
| --- |
| |
| <!-- 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. --> |
|
|
| # all-base |
|
|
| This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on the generator dataset. |
| It achieves the following results on the evaluation set: |
| - Loss: 4.0367 |
|
|
| ## 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 | |
| |:-------------:|:-----:|:-----:|:---------------:| |
| | 6.355 | 0.29 | 500 | 5.3170 | |
| | 5.0354 | 0.58 | 1000 | 4.8924 | |
| | 4.6989 | 0.87 | 1500 | 4.6507 | |
| | 4.4441 | 1.16 | 2000 | 4.5047 | |
| | 4.2822 | 1.45 | 2500 | 4.3873 | |
| | 4.1851 | 1.74 | 3000 | 4.2815 | |
| | 4.0807 | 2.02 | 3500 | 4.2026 | |
| | 3.8813 | 2.31 | 4000 | 4.1635 | |
| | 3.8547 | 2.6 | 4500 | 4.1118 | |
| | 3.8136 | 2.89 | 5000 | 4.0571 | |
| | 3.646 | 3.18 | 5500 | 4.0499 | |
| | 3.5714 | 3.47 | 6000 | 4.0237 | |
| | 3.5592 | 3.76 | 6500 | 3.9907 | |
| | 3.4929 | 4.05 | 7000 | 3.9792 | |
| | 3.3028 | 4.34 | 7500 | 3.9795 | |
| | 3.2966 | 4.63 | 8000 | 3.9665 | |
| | 3.2851 | 4.92 | 8500 | 3.9538 | |
| | 3.1662 | 5.21 | 9000 | 3.9633 | |
| | 3.1138 | 5.49 | 9500 | 3.9624 | |
| | 3.1152 | 5.78 | 10000 | 3.9615 | |
| |
| |
| ### Framework versions |
| |
| - Transformers 4.26.1 |
| - Pytorch 1.11.0+cu113 |
| - Datasets 2.13.0 |
| - Tokenizers 0.13.3 |
| |
