Rate the model
To rate this model, either use /rate or click the buttons below.
1 messages · Page 1 of 1 (latest)
To rate this model, either use /rate or click the buttons below.
It's mid, it doesn't sound like Mori
agreed
Also, the test version of OV2 isn't meant for big datasets
Froggy Simpl said it's only useful for 8-1 mins or 30 secs of audio
and if the dataset is longer than that, then i would recommend training on a regular pretrain
Or if it's insanely clean, use RIN
I hope final versions of OV2 allow any dataset length.
The date set was 30 min and thank you for the feedback for this model. I would try RIN pretraining next time.
No, if your dataset still had some noise, you should have used regular pretrains.
this is the data set, the audio is pretty clean for RIN i think.
Now i understand why your model doesn't sound like Mori.
Bro, don't you realize the audio is still pretty noisy? (It has static noise, room reverb, bg music and low quality voice
And no, if you got an dataset like that, then it's not suitable for RIN.
thank you for the help 👍
bruh
Last update: Mar 3, 2024
@deep sandal We suggest you to read the audio cleanup guides.
Your Mori model got deleted because it didn't sound even a bit like her, had low quality and static noise.
This model has been synced with Weights and is ready to use for free!