v0.8.0

• Major improvements to local ai:
  • now powered by llama.cpp and whisper.cpp
  • in mac it has metal and accelerate support
  • so way fast, windows local ai using gpu in works
• Application size is also now cut by 66%
• Model customization in BYOK AI