<!-- 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. -->
cbt-rarity-all-p8k-new-loop-4-pad
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 4.0928
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: 96
- eval_batch_size: 96
- 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 |
|---|---|---|---|
| 6.2501 | 0.44 | 500 | 5.2101 |
| 4.8858 | 0.87 | 1000 | 4.7710 |
| 4.4914 | 1.31 | 1500 | 4.5271 |
| 4.2765 | 1.75 | 2000 | 4.3635 |
| 4.0636 | 2.19 | 2500 | 4.2561 |
| 3.9151 | 2.62 | 3000 | 4.1691 |
| 3.8177 | 3.06 | 3500 | 4.1096 |
| 3.6143 | 3.5 | 4000 | 4.0756 |
| 3.5995 | 3.94 | 4500 | 4.0220 |
| 3.3792 | 4.37 | 5000 | 4.0261 |
| 3.3439 | 4.81 | 5500 | 4.0032 |
| 3.2316 | 5.25 | 6000 | 4.0074 |
| 3.1585 | 5.69 | 6500 | 4.0048 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
