NamIntegrationTests

Integration tests for the Neural Amp Modeler trainer and core repos.

Repositories

• Trainer — sdatkinson/neural-amp-modeler — Machine learning code for creating new models
• Core — sdatkinson/NeuralAmpModelerCore — Low-level DSP code for real-time model playback

Setup

Clone the trainer and core as siblings of this repo:

parent_dir/
├── NamIntegrationTests/
├── neural-amp-modeler/
└── NeuralAmpModelerCore/

1. Trainer (required): Install neural-amp-modeler, e.g. pip install -e ../neural-amp-modeler, or add it to PYTHONPATH when running tests.
2. Core (required for loadmodel tests): Build the core's loadmodel tool:

   cd ../NeuralAmpModelerCore
   cmake -B build && cmake --build build

3. This project: pip install -e .

Running tests

pytest test/

If the trainer is not installed via pip, add it to PYTHONPATH:

PYTHONPATH=../neural-amp-modeler pytest test/

Tests that require the core's loadmodel tool are skipped if NeuralAmpModelerCore is missing or loadmodel is not built.