jailbreakDetector-v6-onnx

This model is the onnx transformed version of markush1/jailbreakDetector-v6.

Usage

Use with pipeline

from optimum.onnxruntime import ORTModelForSequenceClassification
from transformers import pipeline, AutoTokenizer

model_name = "markush1/jailbreakDetector-v6-onnx"

model = ORTModelForSequenceClassification.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)
classifier = pipeline("text-classification", model=model, tokenizer=tokenizer)
classifier("YOUR_INPUT_GOES_HERE", **{'truncation':True,'max_length':512})

Model description

This version of markush1/jailbreakDetector-v6 is aimed for optimized deployments, especially for execution in environments that lack GPU.

Evaluation on CPU

1.59 s ± 14 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

AMD Ryzen Threadripper 1920X 12-Core Processor, 64 GB RAM