netspec 0.2.5

Python Boilerplate contains all the boilerplate you need to create a Python package.

netspec

netspec allows for the use of neural net emulators of astrophysical photon / particle emission spectra to be trained and then fitted within 3ML to astrophysical spectral data. It is built off pytorch and uses pytorch-lightning as the training interface.

The network structure is adaptable and should be tuned to the need of the simulation. Training data are derived from the outputs of ronswanson and utilities are provided which pre-process the simulation output into suitable spaces for efficient training. Once trained, models can be loaded in as astromodels spectral function as used as any other model for spectral analysis.

• Free software: GNU General Public License v3
• Documentation: https://netspec.readthedocs.io.