MODELS DIRECTORY

On first launch choose mode 1.
The trader will download the default fast GGUF model automatically:
  repo: unsloth/Qwen3.5-0.8B-GGUF
  file: Qwen3.5-0.8B-Q4_K_M.gguf

The larger 2B test model is also available locally:
  file: Qwen3.5-2B-Q4_K_M.gguf
  set SHTENCO_TURBO_MODEL=C:\full\path\models\Qwen3.5-2B-Q4_K_M.gguf

You can use another GGUF without editing the Python file:
  set SHTENCO_TURBO_MODEL=C:\full\path\model.gguf

You can point to a custom llama-server.exe:
  set LLAMA_SERVER=C:\full\path\llama-server.exe

GPU layers default to 999 (offload as much as the llama.cpp build/hardware allows):
  set SHTENCO_TURBO_GPU_LAYERS=999
