Text Classification
Transformers
Safetensors
English
deberta-v2
prompt-injection
prompt-injection-detection
llm-security
llm-safety
ai-safety
deberta
Eval Results (legacy)
text-embeddings-inference
Instructions to use JHC04567/spid-deberta-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use JHC04567/spid-deberta-base with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="JHC04567/spid-deberta-base")# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("JHC04567/spid-deberta-base") model = AutoModelForSequenceClassification.from_pretrained("JHC04567/spid-deberta-base", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -131,7 +131,7 @@ The classifier mode favors precision to avoid blocking legitimate requests. The
|
|
| 131 |
- Epochs: 3, effective batch size 16, max length 256
|
| 132 |
- Calibration: Temperature scaling (T=0.8) on held-out set
|
| 133 |
|
| 134 |
-
**Recommended inference settings:** threshold 0.85, temperature 0.8.
|
| 135 |
|
| 136 |
## Limitations
|
| 137 |
|
|
|
|
| 131 |
- Epochs: 3, effective batch size 16, max length 256
|
| 132 |
- Calibration: Temperature scaling (T=0.8) on held-out set
|
| 133 |
|
| 134 |
+
**Recommended inference settings:** threshold 0.85 (high precision) or 0.80 (catches borderline attacks like DAN-style jailbreaks), temperature 0.8.
|
| 135 |
|
| 136 |
## Limitations
|
| 137 |
|