Realtime Text-to-Speech with GPT-SoVITS
18:43
Пікірлер
@makinganoise6028
@makinganoise6028 2 сағат бұрын
Don't dismiss these apps, my Son started on one on an Android tablet, is now very good on Piano, plays pro on keys and guitar, but don't fool yourself, music is a talent, even for people with perfect pitch and talent it takes thousands of hours, but these apps are a much easier entry than struggling with sheet music to begin with.
@Jarods_Journey
@Jarods_Journey 2 сағат бұрын
Totally on board, the barrier of entry is heavily lowered with tools like this and may act as a gateway to more "traditional" methods. I've found piano vision awesome over this past year!
@johnnyDepp-z5c
@johnnyDepp-z5c 12 сағат бұрын
It's not working
@ferael0013
@ferael0013 Күн бұрын
eh for some reason this does not work for me. the merge button seems to do nothing
@yesyes-om1po
@yesyes-om1po Күн бұрын
you can ask it to make sound effects too, like gunshots or explosions, this is definitely not your average text to speech, this thing is a complete audio AI
@xsploit
@xsploit Күн бұрын
ollama function calling seems alright in my tests
@bigdaddy5303
@bigdaddy5303 Күн бұрын
The most impressive thing is just how well it does accents, notably rhe australian accent. Every other voice cloner ive tried just ends up sounding american. This nails every accent.
@font_net
@font_net Күн бұрын
I want a model that can be given my own data, such as voice and text, and learn based on that data. I want to build a model for Persian. Which of the models you introduced in the video is suitable for building a Persian TTS model?😃😃
@tylerchambliss8379
@tylerchambliss8379 2 күн бұрын
Is there by chance a web UI for any of these just on it's own?
@MyutiConx
@MyutiConx 2 күн бұрын
Is rtx 3070 8gb possible jarods?
@basharalassad1073
@basharalassad1073 2 күн бұрын
the real time voice change models dont seem to work
@mynameissongohanandamnotah17
@mynameissongohanandamnotah17 2 күн бұрын
still okay mine like delay 10 second and just to cut off my half dialogue
@harrietanalytics
@harrietanalytics 3 күн бұрын
thanks for the great tutorial! just wondering do you need to have a local computer with gpu? may we know what type of computer are you having for all these cool AI stuff? Thanks Jarods for the great stuff! 👍
@HoppeTheZ
@HoppeTheZ 3 күн бұрын
After mine was done, I didn't pop up with the interface of the program, anyone else have the same issue? okay update I fixed it but shit is not working, all I hear is some lagging voice LMAO
@chichichichichichiOwO
@chichichichichichiOwO 3 күн бұрын
2:28 what if its none of those? I have a pretty high spec pc originally build for game making (im a programmer) so I know its not my pc but I also tested it on the models that they give you and it still cuts off or mispronounces stuff. the robotics part im able to fix but not anything else.
@飛天柴犬
@飛天柴犬 3 күн бұрын
I think the sentence level breakdown is good to go
@saifayson7303
@saifayson7303 3 күн бұрын
how can i not hear myself while having the voicechanger
@shashwataditya6685
@shashwataditya6685 4 күн бұрын
Which is the best voice cloner F5TTS? Eleven Labs? Or any other? in your opinion
@Headhunternn
@Headhunternn 4 күн бұрын
hi. how to export the A.I. voice as an audio file..?
@zangezaban
@zangezaban 4 күн бұрын
Please make more videos about effcient ways to learn languages. is there a software with the same features as Asbplayer that could run locally?
@justcrap3703
@justcrap3703 4 күн бұрын
I'm at the f5-tts_finetune-gradio part. When I input that in cmd, I get a message: "OSError: [WinError 126] The specified module could not be found. Error loading "C:\Users\(my username)\OneDrive\Desktop\F5-TTS\venv\Lib\site-packages\torch\lib\fbgemm.dll" or one of its dependencies. How should I proceed?
@cleon9976
@cleon9976 4 күн бұрын
can i run this if i save it on an a external ssd
@UpInHeaven
@UpInHeaven 4 күн бұрын
great i can sound like dio now
@jacklamat6315
@jacklamat6315 4 күн бұрын
im using a 4060 with ai
@AL_Momen_F
@AL_Momen_F 4 күн бұрын
I don't know what to say, but my device's specs are four or five times lower than yours. To train it on two hours of recordings with 300 epochs, it took me 40 hours. If I had trained it on 10 hours, it might have taken me a month.
@digitalasylum369
@digitalasylum369 5 күн бұрын
I love Anything LLM for RAG!
@sinayagubi8805
@sinayagubi8805 5 күн бұрын
I hate that I can't bookmark shorts 🩳 Will you post updates like this in long format videos too?
@jayare7750
@jayare7750 6 күн бұрын
is there a way to merge models? I trained 2 different models and figured it's better to have done 1