How to Train and Clone Voice With Accent (workflow using audio webui and OnlySpeakTTS)

Поділитися
Вставка
  • Опубліковано 14 лис 2024

КОМЕНТАРІ • 49

  • @gkhndnc
    @gkhndnc Рік тому +2

    Thank you bro. You make quite high quality content. I'm constantly following and my notifications are on. One thing I'm curious about is what are your pc system specs. Can you give minimum hardware information when describing such artificial intelligence models? Even someone who knows basic code (like me) can do it without any problems. This is because of you, you explain it quite simply and simply. Thanks again.

    • @Natlamir
      @Natlamir  Рік тому +1

      thank you! i have nvidia rtx 3060 with 12GB dedicated ram. i7 with 32GB ram, and running on SSD.

  • @ThiagoGoettems
    @ThiagoGoettems 11 місяців тому

    Have you tried Mangio-RVC?

    • @Natlamir
      @Natlamir  11 місяців тому

      havent tried that one yet

  • @AhmadAli-xv4vd
    @AhmadAli-xv4vd Рік тому

    Thank you for all the efforts your are making... Loving your channel more and more

  • @BilkulBhaiBilkul
    @BilkulBhaiBilkul Рік тому +1

    Hello! I've been trying to run LLaVA locally but for a folder containing Millions of images and save caption in a folder. I have a lot of GPUs but the machine is windows only. Any help or direction would be appreciated!! Thanks!

    • @Natlamir
      @Natlamir  Рік тому +1

      are you using the method from a previous video or some other method? the method i used was through the UI where you can input 1 image at a time. I have not tried programmatically running millions of images from a folder: ua-cam.com/video/ovAzKGaa_og/v-deo.html

    • @BilkulBhaiBilkul
      @BilkulBhaiBilkul Рік тому

      I tried your version and that's only one that works on Windows haha but multiple images would be REALLY GREAT!

  • @huwhitememes
    @huwhitememes Рік тому

    The run.bat file still doesn't work after editing with note ++ to lead to anaconda3 scripts folder. Would you mind me asking, what could I be doing wrong?

    • @Natlamir
      @Natlamir  Рік тому

      can you check what it says when you run the command "where conda"? like for me this is what it says:
      (base) c:\ai>where conda
      C:\Users
      oot\anaconda3\Library\bin\conda. bat
      C:\Users
      oot\anaconda3\Scripts\conda. exe
      C:\Users
      oot\anaconda3\condabin\conda. bat
      line 2 is the scripts path and i am able to use that path in the batch file

  • @thebigbigdaddy
    @thebigbigdaddy 11 місяців тому

    Love it! Can you create a voice assistant to take phone calls via Twilio and GPT?

    • @Natlamir
      @Natlamir  2 місяці тому

      Creating a voice assistant with Twilio and GPT is possible. It would require integrating these technologies along with text-to-speech and speech recognition systems.

  • @LeZappingDuPeuple
    @LeZappingDuPeuple Рік тому +1

    Thanks man 👍

  • @MS-lb9bn
    @MS-lb9bn 10 місяців тому

    I can't train anything. I keep getting "Resampling and then splitting audios into chunks.
    Processing I Lost Something Once.... - Spongebob.wav
    Exception Failed to load audio: [WinError 2] FileNotFoundError" on webui's training page after pressing Resample and split dataset and I don't know why.
    By the way, none of the utils features are working. They keep sayin "error" in red box on webui page.

    • @Natlamir
      @Natlamir  2 місяці тому

      Ensure all audio files are in the correct directory and properly named. Check file permissions and try running the script with administrator privileges if on Windows.

  • @MrDanINSANE
    @MrDanINSANE Рік тому

    Thanks for sharing! your content is very easy to follow 💙
    Is there a similar clone voice which supports Hebrew?

    • @Natlamir
      @Natlamir  Рік тому

      thanks! i will look into that.

  • @SpaceIceDeutschland
    @SpaceIceDeutschland Рік тому +3

    please get a different standrd voice, its just not pleasing at all

    • @Natlamir
      @Natlamir  2 місяці тому

      Thanks for the feedback. I'll explore using different voices in upcoming videos to improve the viewing experience.

  • @CoinHeadlines
    @CoinHeadlines Рік тому

    is there any way we can use openface csv file to make lips snyc

    • @Natlamir
      @Natlamir  Рік тому

      you can use it with DINet to create lip sync

  • @yoann.f
    @yoann.f Рік тому +1

    06:50 : "clip_17.wav" into RVC amplifies the french accent, but the prononciation is all wrong. It's not french.

    • @Natlamir
      @Natlamir  Рік тому +2

      @@Winnetouch777 thanks for letting me know. im not good with noticing subtleties with accents and pronunciations. thanks for letting me know.

  • @the_synapse
    @the_synapse 5 місяців тому

    Cloned voice in french accent of the female english speaker on the last example is not quite the same. It didn't preserve the low pitches of the original voice quite good, seems more like a male voice.

    • @Natlamir
      @Natlamir  2 місяці тому

      Thank you for the detailed feedback. I'll look into improving the voice cloning for low pitches and accents in future updates.

  • @ericanderson5139
    @ericanderson5139 Рік тому

    Does it work for real-time ?

    • @Natlamir
      @Natlamir  2 місяці тому

      Real-time processing depends on your hardware and the specific model used. Some lightweight models can achieve near real-time performance on powerful GPUs.

  • @CoinHeadlines
    @CoinHeadlines Рік тому

    brother DINet & OpenFace is best but its not working showing error i follow all the details you give but there error plz find the easy way of DINET plz thank you

    • @Natlamir
      @Natlamir  2 місяці тому

      For DINet and OpenFace issues, double-check your environment setup and dependencies. I'll consider creating a simplified guide in the future.

  • @mr-s23
    @mr-s23 Рік тому

    not available in other languages ?

    • @stabilitylabs
      @stabilitylabs Рік тому

      waiting for this

    • @Natlamir
      @Natlamir  Рік тому

      the video itself? i currently only make videos spoken in English. thanks.

    • @mr-s23
      @mr-s23 Рік тому

      ​@@Natlamir No, in the video you show how to do it in French, wouldn't it be possible to transform the voice into other languages?

    • @Natlamir
      @Natlamir  Рік тому

      @@mr-s23 it should work with other languahes / accents. you would just need voice samples of the voice you want to clone that is speaking in that language or with that accent so that the RVC model has the same accent when you generate with it.

    • @mr-s23
      @mr-s23 Рік тому

      @@Natlamir Much obliged! I'm going to try it in Portuguese, I'm a big fan of your channel, good luck on your journey!

  • @snuscaboose1942
    @snuscaboose1942 Рік тому

    Nice

    • @Natlamir
      @Natlamir  2 місяці тому

      Thank you! I'm glad you liked the video.

  • @mrGapMan1
    @mrGapMan1 Рік тому +3

    The constant shouting is hilarious.

    • @Natlamir
      @Natlamir  2 місяці тому

      Glad you enjoyed it! The exaggerated expressions were intended to showcase the model's capabilities.

  • @diagorasofmel0s
    @diagorasofmel0s Рік тому

    IDF watching this taking notes

    • @Natlamir
      @Natlamir  2 місяці тому

      The technology has various applications, good to take notes of things.