PSY [ Trained only with "Gangnam Style" ]

# Before Start : > - This version is more of a **"fun version"** of the PSY singer, containing only one specific acapella (Gangnam Style) in the dataset. I'm working on a more "complete" version with more datasets ! > - The result isn't the cleanest possible either, despite the thorough cleaning of the audio. # Last Update : <t:1707084864:R> > - **Model URL :** > - https://huggingface.co/rayzox57/PsyGangnamStyle_RVC/resolve/main/PsyGangnamStyle_v2_Ov2_500e.zip > - **Version :** > - :firev2: RVC V2.0 - Ov2 Super > - **Pitch Extraction Algorithm :** > - RMVPE > - **Epochs - Steps :** > - 500 - 4.0k > - **Dataset :** > - ~ 00:01:29 [ Gangnam Style Only ] > - UVR + Manual Finishing > - **Recommended Usage :** > - Cover > - **Search Feature Ratio :** > - 0.75 > - **Pitch :** > - ( +4 / 0 = Man ) - Depend Base Voice > - You can adjust if you find better version # Previews : > - *I didn't really have any ideas for previews in Korean, so I just put in English previews.* > > - **Preview_Cover_1.wav :** > - *Provide* : Smash Mouth - All Star [ My DIY Acapella - Sample ] > - *Contains External Effects* : Yes > - *Pitch* : +4 > - *Feature Ratio* : 0.75 > > - **Preview_Cover_2.wav :** > - *Provide* : Mustard - Ballin [ My DIY Acapella ] > - *Contains External Effects* : Yes > - *Pitch* : 0 > - *Feature Ratio* : 0.75

Op

Just read that the acapella was done from UVR, this could be done better with the multitrack Gangnam Style stems

Yes, I saw that right after I published this model. After that, I didn't think much of it to see what could be done with it. (In fact, I had the same problem with the S3RL tunes, when I realized afterwards that there are official stems, which are simply 1000x better). When I've got the time, I'll go and have a look (just to train him).

bro this is sick, i'm making a your mine cover with this,

OP