Description

Trained on 1 minute of data ripped from the game wiki (rx11 de-ess de-crackle). there was no mouth clicking or microphone popping on the dataset It's pretty mid at singing, so i recommend using the model only for speaking. Pitch extraction: RMVPE Steps: 2.25k Batch size: 4 Pretrain: Original v2 / 32k Precision: FP32 VA: Martin Schjler Huggingface: Weights.gg: Don't forget to credit my YT channel if used: @fitzyrvc2024; i would appreciate it.