Commit History

  1. Lazy-load CUDA quantizer configuration inside GPU call

    Verified signature
    22338c1
  2. Add bounded bitsandbytes INT8 and NF4 controls

    Verified signature
    1a6347c
  3. Add bounded ZeroGPU BF16 CUDA control

    Verified signature
    bc3f111
  4. initial commit

    Verified signature
    695f3b6