bert-tiny-privacy
This model is a fine-tuned version of prajjwal1/bert-tiny on the beki/privy dataset. It achieves the following results on the evaluation set:
- Loss: 0.0235
Model description
This model can be used to detect personal information traces from JSON, SQL, HTML and XML and can be used as a model for redacting such information.
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: 4e-05
- train_batch_size: 32
- eval_batch_size: 128
- seed: 13434865
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.01
- training_steps: 15000
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.1891 | 0.19 | 2500 | 0.1369 |
| 0.0869 | 0.38 | 5000 | 0.0503 |
| 0.0609 | 0.57 | 7500 | 0.0314 |
| 0.0512 | 0.76 | 10000 | 0.0259 |
| 0.0493 | 0.95 | 12500 | 0.0240 |
| 0.048 | 1.14 | 15000 | 0.0237 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.0
- Tokenizers 0.15.0