ml-stat-Sustech/TorchCP
A PyTorch-based Python library implementing conformal prediction methods for uncertainty quantification in deep learning models.
Not currently ranked — collecting fresh signals.
star history
TorchCP provides researchers with conformal prediction tools built on PyTorch for generating statistically rigorous prediction sets with guaranteed coverage on unseen data. It supports classification, regression, and graph-structured data tasks, and includes adapters for uncertainty quantification in large language models.
Frequently asked
- What is ml-stat-Sustech/TorchCP?
- A PyTorch-based Python library implementing conformal prediction methods for uncertainty quantification in deep learning models.
- Is TorchCP open source?
- Yes — ml-stat-Sustech/TorchCP is open source, released under the LGPL-3.0 license.
- What language is TorchCP written in?
- ml-stat-Sustech/TorchCP is primarily written in Python.
- How popular is TorchCP?
- ml-stat-Sustech/TorchCP has 472 stars on GitHub.
- Where can I find TorchCP?
- ml-stat-Sustech/TorchCP is on GitHub at https://github.com/ml-stat-Sustech/TorchCP.