<!-- 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-mod-datasets3-rarity-all
This model is a fine-tuned version of gpt2 on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 4.3779
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 |
|---|---|---|---|
| 6.7201 | 0.3 | 500 | 5.6554 |
| 5.3777 | 0.6 | 1000 | 5.2100 |
| 5.0257 | 0.91 | 1500 | 4.9662 |
| 4.7428 | 1.21 | 2000 | 4.8246 |
| 4.5916 | 1.51 | 2500 | 4.6972 |
| 4.4886 | 1.81 | 3000 | 4.5927 |
| 4.3213 | 2.12 | 3500 | 4.5355 |
| 4.173 | 2.42 | 4000 | 4.4840 |
| 4.1402 | 2.72 | 4500 | 4.4195 |
| 4.0833 | 3.02 | 5000 | 4.3844 |
| 3.8496 | 3.33 | 5500 | 4.3743 |
| 3.8398 | 3.63 | 6000 | 4.3421 |
| 3.8193 | 3.93 | 6500 | 4.3113 |
| 3.6103 | 4.23 | 7000 | 4.3294 |
| 3.5592 | 4.53 | 7500 | 4.3199 |
| 3.5442 | 4.84 | 8000 | 4.3041 |
| 3.4575 | 5.14 | 8500 | 4.3158 |
| 3.3572 | 5.44 | 9000 | 4.3191 |
| 3.3595 | 5.74 | 9500 | 4.3171 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.11.0+cu113
- Datasets 2.13.0
- Tokenizers 0.13.3
