I'll show you how to generate free text-to-speech or clone voices from samples using only AI and your computer, for free!
✅ Hostinger $1.79 ➜ [ Ссылка ]
CODE for extra 10% OFF ➜ EMITREVIEWS
Grab some cheap hosting for your next project!
[April 26th UPDATE. This tutorial might result in errors after Python version update, I've released an updated version that 100% works — [ Ссылка ]]
For the free method to work, you'll need to have an Nvidia graphics card that has Cuda Cores (because you'll be using your graphics card to render out the voices)
Here's all of the software you'll need
1. Original creator page — [ Ссылка ]
2. Anaconda — [ Ссылка ]
3. PyTorch — [ Ссылка ]
4. Git — [ Ссылка ]
5. Audacity — [ Ссылка ]
To install tortoise:
(in the pinned comment)
To install pysoundfile:
conda install -c conda-forge pysoundfile
To generate AI speech:
python tortoise/do_tts.py --text "I'm going to speak this" --voice random --preset fast
Change the text within " ", this will be words that will be generated. Change the voice to whichever voice you want to use for the generation. I recommend using --preset fast for a first try, if you like the results, you can re-generate using --preset standard for better quality.
If your Anaconda Promt doesn't navigate to the correct directory automatically use:
cd tortoise-tts
I recommend checking out some of the paid generation options for a faster way that doesn't rely on you having an Nvidia graphics card. There are plenty to choose from but I went for [ Ссылка ]
You can use this software to create "text-to-speech" articles. Meaning users would be able to "listen" to an article on your website instead of reading.
