This is really great, can this be called through the api

[XXXXRT666]

https://github.com/RVC-Boss/GPT-SoVITS/blob/main/api.py

[Separatee]

What the previous person said is correct. But you can also switch to fast_inference if you want a more powerful api. Switching to this branch is very simple! You can switch to this branch while keeping your master branch intact by using the

git clone --depth=1 -b fast_inference_ https://github.com/RVC-Boss/GPT-SoVITS.git fast_inference

Then overwrite the fast_inference directory with all files except .git and docker into the GPT-SoVITS main directory Then use pip the way you did (if you use All-in-one Packages, you need to tweak the command)

pip install wordsegment

[aqpmzngldh]

What the previous person said is correct. But you can also switch to fast_inference if you want a more powerful api. Switching to this branch is very simple! You can switch to this branch while keeping your master branch intact by using the

git clone --depth=1 -b fast_inference_ https://github.com/RVC-Boss/GPT-SoVITS.git fast_inference

Then overwrite the fast_inference directory with all files except .git and docker into the GPT-SoVITS main directory Then use pip the way you did (if you use All-in-one Packages, you need to tweak the command)

pip install wordsegment

I don't know much about this, I wonder if this project can pass text through api after training the model according to a model, and then respond to the voice link to me？

[Separatee]

link

@aqpmzngldh truely it is. but not a link, it will be wav file directly with Internet code 200. for more details, you can switch to fastinference github branch and see the api_v3.py in this file, it described what you d like to know(but need translate tools) see this

[Separatee]

For convenience, I translated the Api.v3.py doc. pure luv from zh-cn Trans.txt