Although we have enhanced the sampling in the original Version 1 to suppress electronic sounds, there are still electronic sounds occurring in certain ranges. General conversation or TTS usage is relatively free, but if you wish to use it for high-pitched song covers, it is recommended to use it excluding the Index. We have determined that it is pointless to look for more of Karina's recording sources, so the production of Karina will be temporarily suspended until new audio sources are obtained.

Information :

Model :
Model Name : Karina (카리나)
Job : Kpop Idol

Train info :
RVC 2
1200 Epochs
Batch size per GPU : 40

Dataset - source Info :
12 mins of speech
20 mins of Rebuilt Speech source
4 mins of Sing

Dataset - source correction :
Volume Normalizing (V1)
Basic de-noise (V1)
De-Noise and dehummers (V2)
Audio Pitch Range (Vocal) Reinforcement (V2)

Model Link - https://huggingface.co/SeoulStreamingStation/IU-Voice-MultiLanguage-V1/resolve/main/Karina_V2.zip?download=true

Additional Details:

Tags: No tags available

Download Link: https://huggingface.co/SeoulStreamingStation/IU-Voice-MultiLanguage-V1/resolve/main/Karina_V2.zip?download=true