Neural Amp Modeler is a Python framework for training neural network models that emulate guitar amplifiers.
The tool addresses the problem of capturing the complex, nonlinear behavior of guitar amplifiers through machine learning. It trains neural networks on input-output recordings of amplifiers and exports the resulting models to a standardized .nam file format. This approach allows the acoustic characteristics of physical amplifiers to be digitally reproduced without requiring hand-crafted signal processing chains.
The project suits musicians and audio engineers who want to create digital emulations of specific amplifiers or design custom amp models. It is designed as a training and export pipeline; real-time playback of trained models happens in a separate companion plugin application. The tool is part of a broader ecosystem that includes both the training framework and the plugin for using exported models in production audio workflows.
Development activity shows consistent engagement with the codebase. The project maintains organized documentation hosted online with build instructions for multiple platforms. The repository structure separates concerns between model training and inference, with clear pointers to the companion plugin repository for users interested in the complete workflow.
