NEEDED TO RUN:
Install ollama
Install the llama3 model ( in cmd after ollama is installed >ollama run llama3)
Using the ModelFile construct your own "Riku" Ai see the ollama git repo to learn the commands

install nivida cuda toolkit
GPU execution requires the following NVIDIA libraries to be installed:
cuBLAS for CUDA 11
cuDNN 8 for CUDA 11

Install ffmpeg for text to speech: https://www.youtube.com/watch?v=jZLqNocSQDM
The SystemEngine works out of the box using your system's built-in TTS capabilities. No additional setup is needed.

Install the required libraries and packages
To run the code use: python app.py

FOR YOLOv10:
https://youtu.be/D2HKry0k7uU?si=aFrNV2p3S_t2HiVy

FOR THE VIRTUAL BODY:
If using Vtube studio avatar, install vtubestudio and configure it enableing api. vtubestudio can be installed on steam.
Download and run the VTS Desktop Audio application and configure it so that the animations in vtubestudio are more acurate.
Once complete, Configure the avatar on vtubestudio by changing the "Mouth Open" Configuration. Select the Input button and scroll until you find the option DesktopAudioVolume, select this option. The avatar should now be configured to use the desktop audio for automated mouth movements as long as VTS Desktop Audio is open and running.
If you are trying to have the output of the AI's voice to not be your default device so that you can prevent other noise. Use a virtual audio cable such as VB-Audio and for extra configurations of audio devices use the software Voicemeeter. Then configure the default device settings on ffmpeg and the various dependecies, this can be done in windows audio settings on your system.

References:
https://github.com/SYSTRAN/faster-whisper
https://github.com/ollama/ollama
https://github.com/KoljaB/AIVoiceChat
https://github.com/KoljaB/RealtimeTTS
https://lualucky.itch.io/vts-desktop-audio-plugin
https://vb-audio.com
https://voicemeeter.com/?gad_source=1&gclid=CjwKCAjwz42xBhB9EiwA48pT7xGu-Xz0A-K4SwhsPbtFOLoVj2aTquKYkyegJIx48zvcJDGogxIJLRoCqhcQAvD_BwE
https://denchisoft.com
https://huggingface.co/stabilityai/stable-diffusion-2-1/tree/main