Same here! Extremely excited for this for my sci fi novel series. It’s too bad I only have an ipad pro and can’t download the git. Just out of random curiosity, what is your novel about?
Ive actually started working on a version of this that can run in google colab that can use gigantic text files as input and separates them into sentences before combining the wav files idk
made a notebook that can take gigantic audiofiles should fix the short text limiting issue the vid has a link in it: ua-cam.com/video/qXFM7g7la9U/v-deo.html
Please make a video on how to clone our voice and generate longer Speech for free. Plz create video on this, you previous video on voice clone is good but we cant generate longer speech through it.
Just use a python NLP package to split a long input into individual sentences as a string list, then have it run through every string in the list- generating the wav output file for each one and then merge all the wav files together.
@@drewthomasson949 (a) that'll sound like crap, and (b) it's not cloning but merely a more efficient way of recording and replaying. Totally different thing.
Maybe they didn’t design or train the system so the second part of code switching would have accents. Maybe it was an emerging property due to bias in the training data… i.e. people who speak two languages are likely going to be speaking the mother tongue first and then have an accent derived from their mother tongue in the foreign language.
I installed from github. It didn't take so long to install but it says that it needs 40GB of VRAM to run. So I wanted to run it on the CPU. But I'm looking everywhere and no idea how to run it on CPU. I tried it on google colab and it's working but it is slow indeed. if I could run it on my CPU I wouldn't really care if it's slow if it can do the work.
@@The_Daily_Meow I tried to install local on mac. The code is not producing any voice. And this author removed my comments after 2 hours so I have to reply here. Did you run this with colab? It seems need to buy some GPU services to run. I tried locally it is not using GPU so it is very slow and no voice.
Given the tags and the music note token, my guess is that this was trained on closed-captioned films and Tv shows
This is EXACTLY what I've been looking for to narrate my novel, turning it into an audio book!
Same here! Extremely excited for this for my sci fi novel series. It’s too bad I only have an ipad pro and can’t download the git. Just out of random curiosity, what is your novel about?
Ive actually started working on a version of this that can run in google colab that can use gigantic text files as input and separates them into sentences before combining the wav files idk
made a notebook that can take gigantic audiofiles should fix the short text limiting issue the vid has a link in it: ua-cam.com/video/qXFM7g7la9U/v-deo.html
For now this model i very limited and not really reliable, two sentences only
@@Zale370 Yeah, it's in a beta phase. But, it's promising!
This model is good for very short prompts, if you put something larger it just stops after two paragraphs.
Yeah it's cool, right now a novelty to play with but not practical to use.
I'd like a tutorial for cloning my voice with Bark.
Could you do a toturial for that?
Yes!
The emotions and inflections are pretty great. However, the voices themselves still sound pretty synthetic. Pretty good for a free model though.
Agree!
hey,mate... are you taking bark studio waitlist ??
Thanks for video you are killing it bro ...Any update on how to create AI avatar free..?
I was almost wondering it the narrator of this video had AI generated speech. 😀
Is there a code to run the length of audio to 1 minute as currently only 11 s audio is possible
Can you do that in an iPhone?
Bro., can you please try this app with long text.
Are there any tools out there that can toggle the audio speed?
where are you from....? Just want to know.
can we use this voice for you tube..?
It's actually decent, even if there is some robotic voice.
Superb sir
Superb ❤
Can it train my voice to speak Chinease ?
I think we listen to the same channels! 🤣
Nouse in the audio, it's free but not really clean
AI are LOve :)
3:58 I am French and I understood but these are really not exceptional, it's like someone with a big American accent speaking in French
Seems like it's turned for English at the moment. Hopefully they have more models on the way
Sadly, we can't use this to clone our own voice.
so far
It better not be bark 😂
Is bark bad?
@@IForgetWhatISay bark has training limitations.
Training is limited to pre written text.
Please don't show how to use bark with somebody else's voice, it's not safe!
Yeah, dont forget to never feed it after midnight too.
I am not able to use long text in it
We would need
1) a way to train the model with our own voice
2) a way to clear the audio, it has robotic sounds
8:20 looking forward for that video about voice cloning with bark
Please make a video on how to clone our voice and generate longer Speech for free. Plz create video on this, you previous video on voice clone is good but we cant generate longer speech through it.
Just use a python NLP package to split a long input into individual sentences as a string list, then have it run through every string in the list- generating the wav output file for each one and then merge all the wav files together.
@@drewthomasson949 🤯
@@drewthomasson949
Dude please, shut up.
@@drewthomasson949 (a) that'll sound like crap, and (b) it's not cloning but merely a more efficient way of recording and replaying. Totally different thing.
The French sample wasn't too bad, but sounded generated.
While the English ones are simply outstanding.
quedo a la espera del video de claonacion de voz con esta hermosa libreria... como siempre... excelente video
Great video. Looking forward to the method to create our own voices.
can you make a video explaining how to actually run it on the computer? Think I have the install.
Sure, will look into it.
in short which is the best text to audio currently when you can make audio with emotions and more.
how to install step by step. make a video
how can I use bark to voice clone? I can not find the vedio.
Maybe they didn’t design or train the system so the second part of code switching would have accents. Maybe it was an emerging property due to bias in the training data… i.e. people who speak two languages are likely going to be speaking the mother tongue first and then have an accent derived from their mother tongue in the foreign language.
That's a possibility
how can I clone my own voice?
THANKS how can we create singing voice or track
I tried it out a few days ago.
In order to generate 1 sentence on my Intel 10700 It would take 1 hour ! :)
try again it is now 10 to 100 times faster with CPU! But still slow compared to GPU.
@@DrHanes Thx. Ill have a look.
how to run it on CPU?
I installed from github. It didn't take so long to install but it says that it needs 40GB of VRAM to run. So I wanted to run it on the CPU. But I'm looking everywhere and no idea how to run it on CPU. I tried it on google colab and it's working but it is slow indeed. if I could run it on my CPU I wouldn't really care if it's slow if it can do the work.
@@The_Daily_Meow I tried to install local on mac. The code is not producing any voice. And this author removed my comments after 2 hours so I have to reply here. Did you run this with colab? It seems need to buy some GPU services to run. I tried locally it is not using GPU so it is very slow and no voice.
can you share the colab link? thanks
awesome!!