llamafile is not really “effective”. it’s incredibly impressive, but it’s the opposite of effective. it’s a collection of a bunch of hacks reliant on coincidences in OS design, and works by basically recompiling itself on the fly to work with different architectures.
if you want effective, run llama.cpp compiled with actual optimizations for your platform.
it’s a good idea to not look to deeply into the historic actions of the creator of llamafile. she’s pretty polarising.
I don’t care who they are or what their Xitter history is.
The tools is great, the tool is not backdoored. I ruthlessly use effective tools that I can get my hands on.
Using open source software on its own does not even entails economic support for its creator.
llamafile is not really “effective”. it’s incredibly impressive, but it’s the opposite of effective. it’s a collection of a bunch of hacks reliant on coincidences in OS design, and works by basically recompiling itself on the fly to work with different architectures.
if you want effective, run llama.cpp compiled with actual optimizations for your platform.
Care to share some highlights?