---
library_name: transformers
license: mit
base_model: FacebookAI/roberta-base
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: roberta-base-trustairlab-jailbreak-augmented
  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. -->

# roberta-base-trustairlab-jailbreak-augmented

This model is a fine-tuned version of [FacebookAI/roberta-base](https://huggingface.co/FacebookAI/roberta-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1497
- Accuracy: 0.9564

## 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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 10

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Accuracy |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
| 0.7089        | 1.0   | 4240  | 0.2427          | 0.9236   |
| 0.2055        | 2.0   | 8480  | 0.1981          | 0.9365   |
| 0.1062        | 3.0   | 12720 | 0.1940          | 0.9459   |
| 0.2229        | 4.0   | 16960 | 0.1855          | 0.9521   |
| 0.3277        | 5.0   | 21200 | 0.1498          | 0.9559   |
| 0.2119        | 6.0   | 25440 | 0.1607          | 0.9595   |
| 0.0957        | 7.0   | 29680 | 0.1673          | 0.9591   |
| 0.0421        | 8.0   | 33920 | 0.1794          | 0.9594   |


### Framework versions

- Transformers 5.2.0
- Pytorch 2.10.0+cu128
- Datasets 4.5.0
- Tokenizers 0.22.2