|
--- |
|
license: mit |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: gelectra-base-injection-pt_v1 |
|
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. --> |
|
|
|
# gelectra-base-injection-pt_v1 |
|
|
|
DEPRECATED - PLEASE USE NEWER GELECTRA OR DEBERTA VERSION |
|
|
|
This model is a fine-tuned version of [deepset/gelectra-base](https://huggingface.co/deepset/gelectra-base) on a closed prompt injection dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0163 |
|
- Accuracy: 1.0 |
|
|
|
## Model description |
|
|
|
The model classifies prompts as injections or legitimate questions. |
|
|
|
## 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: 5e-05 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 3 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:| |
|
| No log | 1.0 | 45 | 0.2042 | 0.9211 | |
|
| No log | 2.0 | 90 | 0.0247 | 1.0 | |
|
| No log | 3.0 | 135 | 0.0163 | 1.0 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.27.4 |
|
- Pytorch 2.0.0+cu118 |
|
- Tokenizers 0.13.3 |
|
|
