The Dactyl AI-Generated Text Detector in ONNX format is a machine learning model designed to identify AI-generated text. This repository provides a format conversion of the original DACTYL model, enabling inference on CPU using onnxruntime without altering the model's weights or behavior. The model is based on microsoft/deberta-v3-large and has been fine-tuned with Empirical X-Risk for the task of text classification, specifically targeting the detection of AI-generated content.
The ONNX build allows users to run the model outside its original PyTorch environment, facilitating integration into workflows that rely on ONNX-compatible runtimes. The repository is intended for those who require an AI-generated text detection model in a more portable and interoperable format. The conversion retains the original model's capabilities for text classification and AI-generated text detection, as described in the referenced arXiv paper.
This tool is distributed under the MIT License, inheriting the licensing terms from the original model. Users are asked to credit and cite the original authors when utilizing the model. The repository does not introduce new features beyond the format conversion and is not a standalone application but rather a model artifact for integration into broader systems or research workflows.
In the Other AI space, Dactyl Ai Text Detector takes a focused approach. It focuses on identifying and classifying AI-generated text in English documents for authenticity verification. It is built as an open-source project for AI researchers. Dactyl Ai Text Detector is open source under the MIT license. Dactyl Ai Text Detector is available on the web and the command line.
Behind Dactyl Ai Text Detector is GabeuxDev, and the product first shipped in 2025. Key capabilities include AI-generated text detection, text classification, and ONNX format.
Latest indexed changes and source events
GabeuxDev/dactyl-ai-text-detector-onnx verified by the PulseGate indexer
Other apps tracked under the same category.
