sandslides

10 ms·

1. 1.▲
   by sandslides 7mo ago

   could I just leave my favourite thing ever here? thanks :) https://en.wikipedia.org/wiki/Hilbert%27s_paradox_of_the_Gra...
2. 2.▲
   by sandslides 2y ago

   So Ghostbusters II is now reality? :)
3. 3.▲
   VoiceCraft: Open-source high quality voice cloning / voice editing(github.com)

   5 points by sandslides 2y ago|1 comments
4. 4.▲
   by sandslides 2y ago

   The model weights have been uploaded to Huggingface : https://huggingface.co/pyp1/VoiceCraft This seems to be really high quality judging by the demo's. Not had time to try it for myself Demos : https:/&#x2
5. 5.▲
   by sandslides 3y ago

   The speech generated is the best I've heard from an open source model. The one test I made didn't make an exact clone either but this is still early days. There's likely something not quite right. The cloned voice does speak
6. 6.▲
   by sandslides 3y ago

   The LibriTTS demo clones unseen speakers from a five second or so clip
7. 7.▲
   by sandslides 3y ago

   Yes, I noticed that. Doesn't seem right does it
8. 8.▲
   by sandslides 3y ago

   Not sure. The only inference demos are colab notebooks. The models are approx 700mb each so I imagine it will run on modest gpu
9. 9.▲
   by sandslides 3y ago

   Just tried the collab notebooks. Seems to be very good quality. It also supports voice cloning.
10. 10.▲
    StyleTTS2 – open-source Eleven-Labs-quality Text To Speech(github.com)

    725 points by sandslides 3y ago|234 comments