GuardEx DistilBERT Safety Classifier
Binary classifier that labels a piece of text as safe or unsafe. Used by
GuardEx, an LLM guardrail library, as one of
its content-safety models. This is the fastest of the GuardEx classifiers, with
the lowest accuracy of the three.
Labels
| id | label |
|---|---|
| 0 | safe |
| 1 | unsafe |
Performance
Held-out evaluation:
- F1: 0.720
Usage
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
repo = "AtliQ-Technologies/guardex-distilbert-safety"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForSequenceClassification.from_pretrained(repo)
enc = tok("text to check", return_tensors="pt", truncation=True, max_length=128)
with torch.no_grad():
logits = model(**enc).logits
print(model.config.id2label[int(logits.argmax())]) # "safe" or "unsafe"
Details
Fine-tuned from martin-ha/toxic-comment-model. Max sequence length 128.