RaidenSilver/desktop-pet-AI

Preface

The reason for creating this project is my curiosity and some unknown psychological activities.

Acknowledgements:

LLM: Thanks to the X.ai TTS: Thanks to the GPT-Sovits project PUI: Thanks to the MyFlowingFireflyWife project I’m not sure who exactly to thank… Anyway, I won’t do anything unethical. Thanks to them.

Explanation

I can’t stand the upload mechanism of Github repositories… so I deleted the files… There are a total of 2 parts, and the source code can be found in the following repositories:

TTS PUI

Building from Source:

TTS:

Unzip TTS.zip to TTS (root directory, do not place the entire folder into TTS, about 23 items), then deploy according to the official version. I used Python’s venv virtual environment, not Conda:

python -m venv venv
.\venv\Scripts\activate
pip install -r requirements.txt

Otherwise, you need to modify the batch file. TTS is using the V2 Models. ###You can download the model here: Huggingface

If TTS reports an error, please try the following command:

.\venv\Scripts\activate
python

import nltk
nltk.set_proxy('http://127.0.0.1:10809') #Replace it with your proxy port
nltk.download('cmudict')
nltk.download('averaged_perceptron_tagger')
nltk.download('averaged_perceptron_tagger_eng')

LLM:

I will use Grok's API to obtain LLM support. You can register and obtain your Grok API at X.

IMPORTANT:

Fill in the Grok API key yourself, on line 2 of the .\PUI\config.py file.

PUI:

Unzip PUI.zip to PUI (similarly, do not place the folder, but the files, about 13 items)

python -m venv venv
.\venv\Scripts\activate
pip install -r requirements.txt

This completes the three parts (Maybe?).

Finally, confirm if all models are in place…

Then double-click API.bat in the LLM folder and API.bat in the TTS folder. After both are loaded, open StartWithAPI.bat in the PUI folder. Then you can enjoy it.

STT

This is a local update for converting speech to text. The installation method is as follows:

python -m venv venv
.\venv\Scripts\activate
pip install -r requirements.txt

At the same time, download the model to the root directory.

git clone https://huggingface.co/Systran/faster-whisper-large-v3

Now, you can double-click the Main.batfile to start the STT API service.

Conclusion

This repository is a consolidated archive for the DPAI project. {Github}(https://github.com/Lin-Silver/desktop-pet-AI)

Find the integration package here: huggingface