EGFxSet

该数据集旨在评估神经网络架构在48kHz采样率下复制弹簧混响特性的有效性。其中，60%的样本用于训练，20%用于验证，另外20%则用于评估，这些样本是通过随机分割的方式进行的。该数据集的任务是神经音频效果建模。

This dataset is designed to evaluate the effectiveness of neural network architectures in replicating the characteristics of spring reverb at a 48 kHz sampling rate. Specifically, 60% of the samples are allocated for training, 20% for validation, and the remaining 20% for evaluation, with the dataset split via random partitioning. The task of this dataset is neural audio effect modeling.