Skip to content

Untested prototype

Pre-release
Pre-release

Choose a tag to compare

@GeekMaster117 GeekMaster117 released this 29 Aug 10:45
· 147 commits to main since this release

Fixed minor bugs:

  • llama not found error during benchmark, if memory is not enough.

Improvements:

  • Suppressed warning CUDA_PATH or nvcc.exe not found.

This is an untested version of the prototype

Minimum Requirements:

  • Windows OS
  • 4GB of available RAM

Recommended Requirements:

  • Nvidia GPU 10 series or above
  • 4GB of available VRAM

How to use:

  • Run install.exe
  • Run observe.exe (This monitors app usage)
  • Run reflect.exe (To generate suggestions)