SDXL Base and Refiner 01 - INT8

These are not INT8 ConvRot.

These models are INT8 Tensorwise. if you are on AMD ROCm, these models will pack a real punch if you take the time to custom build ComfyUI with ROCm\FlashAttention + AMD Triton / AITER... But they are expected to pack a performance boost on Nvidia cards too.

Minimum start up parameters for ROCm\FlashAttention + AMD Triton:

IMPORTANT: Do not install flashattention offered by pip get official ROCm\flashattention from github.

export FLASH_ATTENTION_TRITON_AMD_ENABLE=TRUE

python main.py \
  --use-flash-attention \
  --disable-xformers 

More details and benchmarks coming soon.