| license: mit | |
| tags: | |
| - pytorch | |
| - reflectometry | |
| - x-ray-reflectometry | |
| - bayesian-analysis | |
| - neural-posterior-estimation | |
| - scientific-computing | |
| - materials-science | |
| library_name: pytorch | |
| pipeline_tag: other | |
| # PANPE Reflectometry Model Weights | |
| Pre-trained model weights for **Fast and Reliable Probabilistic Reflectometry Inversion with Prior-Amortized Neural Posterior Estimation**. | |
| **Note**: This repository provides model weights for community access. The original work and model training were performed by Vladimir Starostin and colleagues. This is not an official repository by the original authors, but rather a community contribution to make the pre-trained weights easily accessible via HuggingFace. | |
| ## Model Description | |
| This repository contains the trained neural network weights for the PANPE (Prior-Amortized Neural Posterior Estimation) model designed for Bayesian reflectometry analysis. The model enables fast and reliable probabilistic inversion of X-ray reflectometry data. | |
| ### Using the Model | |
| To use these weights, you need the full PANPE package. Get the complete code and installation instructions from: | |
| - **GitHub Repository**: https://github.com/mlcolab/panpe | |
| - **Zenodo Dataset**: https://zenodo.org/records/14267737 | |
| ## Files in this Repository | |
| - `saved_models/model_panpe-2layers-xrr.pt`: Pre-trained PyTorch model weights | |
| - `configs/panpe-2layers-xrr.yaml`: Model configuration file | |
| - `LICENSE.txt`: MIT License | |
| ## License | |
| This project is licensed under the MIT License - see the [LICENSE.txt](LICENSE.txt) file for details. | |