> NTransformer High-efficiency C++/CUDA LLM inference engine. Runs Llama 70B on a single RTX 3090 (24GB VRAM) by streaming model layers through GPU memory via PCIe, with optional NVMe direct I/O that bypasses the CPU entirely.

untested:

https://github.com/xaskasdf/ntransformer

if it's worth i extract the core value out of the "conversation" with myself. could be a page in any knowledge base, a snippet in my ide, or whatever suits well... then the email thread goes either into archive or into data nirvanah.

...sending myself an email

If you are a dark mode addicted like me:

go for NORD theme

https://github.com/EliverLara/Nordic

and I love this icon set (white)

https://www.xfce-look.org/p/1277095

for more NORD integration have a look here:

https://www.nordtheme.com/ports

have fun

I feel u!