Text Classification
Transformers
Safetensors
English
deberta-v2
prompt-injection
security
span-detection
guardrails
ai-safety
agents
llm-security
text-embeddings-inference
Instructions to use Unplug-AI/unplug-tiny-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Unplug-AI/unplug-tiny-v1 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="Unplug-AI/unplug-tiny-v1")# Load model directly from transformers import AutoTokenizer, DebertaV2ForDualHead tokenizer = AutoTokenizer.from_pretrained("Unplug-AI/unplug-tiny-v1") model = DebertaV2ForDualHead.from_pretrained("Unplug-AI/unplug-tiny-v1") - Notebooks
- Google Colab
- Kaggle
link demo space and Guard.with_tiny usage
Browse files
README.md
CHANGED
|
@@ -99,6 +99,9 @@ Strong on Neuralchemy, BIPIA indirect, notinject, public validation. Known weakn
|
|
| 99 |
|
| 100 |
```python
|
| 101 |
from unplug import Guard
|
| 102 |
-
|
|
|
|
| 103 |
result = guard.scan(user_text)
|
| 104 |
```
|
|
|
|
|
|
|
|
|
| 99 |
|
| 100 |
```python
|
| 101 |
from unplug import Guard
|
| 102 |
+
|
| 103 |
+
guard = Guard.with_tiny() # auto-downloads Unplug-AI/unplug-tiny-v1
|
| 104 |
result = guard.scan(user_text)
|
| 105 |
```
|
| 106 |
+
|
| 107 |
+
**Interactive demo:** [Unplug-AI/unplug-tiny-demo](https://huggingface.co/spaces/Unplug-AI/unplug-tiny-demo) (span highlights + redaction).
|