Because, potentially, we haven't told you enough!

[metal3d]

I'd like to thank you for all your hard work and all the tools you've made available to the community. Whether it's the library, the conversion tools, or the quality of the project that has enabled it to be integrated into other languages (such as Python and Go).

Yes, my message is just that, a thank you.

You've made it possible for us to use models that had a hard time running on modest hardware. In particular, my computer, which is over 10 years old and only supports AVX (not AVX2), can finally run LLama3, Mistral or Phi3 models.

You're opening the door to innovation for everyone. I've tested a number of solutions... I can't count the number that have refused to start me up because my PC (which has an RTX 3070 and 32GB of RAM, with an i7 core) doesn't support AVX2. My PC shouldn't have been obsolete for so little. It works, it's on Fedora 40, and it's fast.

And llama.cpp allows me not to let it go to waste. So, really, from the bottom of my heart: Thank you very much.