可调节计算量的神经放大器模型，用时不重新训练。
Slimmable NAM: Neural Amp Models with adjustable runtime computational cost
- 通过可缩放结构实现模型大小与计算成本动态调整
- 无需重训练即可在不同计算开销间切换，性能稳定
- 适合音乐人实时使用，兼顾音色精度与算力需求
本文提出可缩放神经放大器模型（Slimmable NAM），其规模和计算成本可在不进行额外训练且计算开销极小的情况下灵活调整，使音乐人能轻松在模型精度与计算资源之间权衡。该方法在常用基线模型上进行了性能评估，并开发了实时音频效果插件演示，验证了其在实际应用中的可行性与有效性。
原文摘要 · Abstract (English)
This work demonstrates "slimmable Neural Amp Models", whose size and computational cost can be changed without additional training and with negligible computational overhead, enabling musicians to easily trade off between the accuracy and compute of the models they are using. The method's performance is quantified against commonly-used baselines, and a real-time demonstration of the model in an audio effect plug-in is developed.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。