Looks absolutely stellar!

Another step forward for local AI / local LLM inferencing on (very!) resource-constrained consumer GPU hardware!